The Lower/Middle Paleolithic Periodization in Western Europe

� 2006 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/2006/4705-0001$10.00

Current Anthropology Volume 47, Number 5, October 2006 709

The Lower/Middle Paleolithic Periodization inWestern Europe

An Evaluation

by Gilliane F. Monnier

In the late nineteenth century, the European Paleolithic was divided into stages, each of which wascharacterized by a distinct stone tool type known historically as a fossile directeur, or index fossil.Today, these index fossils are no longer explicitly used to date assemblages because they are knownto overlap widely, but they continue to be used as key components in the periodization of thePaleolithic. This study addresses two major questions: (1) How have archaeologists justified retainingthese index fossils to distinguish the Lower from the Middle Paleolithic? and (2) Does the diachronicpatterning of these tool types support this periodization? The results reveal (1) that the overlap ofindex fossils was a known problem from the beginning and prehistorians repeatedly modified theclassification to accommodate data which documented this overlap without rejecting its fundamentalindex-fossil-based core and (2) that, while temporal trends agreeing with the Lower/Middle Paleolithicperiodization can be identified, they are insignificant compared with the chronological variationobserved in each of these tool types. The Lower/Middle Paleolithic periodization should thereforebe revised on the basis of a comprehensive examination of multiple lines of evidence, not just lithictypology.

The past two decades have witnessed intense research anddebate regarding the Middle/Upper Paleolithic transition inEurope and its relationship to the origins of modern humans.There has been little interest, however, in the Lower/MiddlePaleolithic transition (but see, e.g., Ronen 1982a). There areseveral reasons for this lack of interest. First, the Lower/MiddlePaleolithic transition is less dramatic than the later one, itsexact date being unclear and its existence sometimes calledinto question (e.g., Tuffreau, Lamotte, and Marcy 1997). Sec-ond, the chronological resolution from this time period hasbeen extremely poor, although it is improving. Finally, theLower and even the Middle Paleolithic have generally beenregarded as static and unchanging (Kuhn and Hovers 2005).Therefore, while recent volumes have focused on variabilityin the Middle Paleolithic (e.g., Roebroeks and Gamble 1999;Hovers and Kuhn 2005) and on the earliest occupation ofEurope (Bonifay and Vandermeersch 1991; Roebroeks andvan Kolfschoten 1995a), very few have tried to refine ourknowledge of what happened at the interface of the Lowerand the Middle Paleolithic. Yet, a comparison of the two

Gilliane F. Monnier is Instructor and Fellow in the Department ofAnthropology at the University of Minnesota (301 19th Ave. South,Minneapolis, MN 55455, U.S.A. [[email protected]]). The pre-sent paper was submitted 11 V 05 and accepted 6 XII 05.

transitions reveals that they are hampered by similar prob-lems. The first problem is that an archaeological transitiontends to dichotomize the archaeological record, producing a“before” and an “after” which are often seen as internallycohesive. Many years ago, Straus (1983) pointed out that thetreatment of the Middle and the Upper Paleolithic as singleentities has magnified the differences between the two whilereducing the variation within each period. Upon closer ex-amination, it turns out that there are more continuities acrossthe transition than is generally acknowledged (e.g., d’Erricoet al. 1998). Recent studies have also shown that the transitionvaried geographically: in Africa, for example, the Middle/LaterStone Age transition was “gradual and episodic,” occurringover a span of at least 200,000 years (McBrearty and Brooks2000), while in China the Middle/Upper Paleolithic transition(ca. 27,000–30,000 years ago) appears to have been so abruptand change in the period preceding it so gradual that Gaoand Norton (2001) have proposed eliminating the MiddlePaleolithic altogether.

The second problem with these transitions is that we oftenforget that they are artificial boundaries designed to providestructure to a complex record and, rather than being con-ceived of as permanent or real, should be frequently examinedand revised (Roebroeks and Corbey 2001). Yet, the Lower andMiddle Paleolithic index-fossil-based definitions created over

710 Current Anthropology Volume 47, Number 5, October 2006

100 years ago, when the Acheulean was characterized by bi-faces and the Middle Paleolithic by Levallois technology andretouched flake tools, are still defined the same way despitethe vast temporal distribution of each type and huge overlapsbetween them. Representative definitions are as follows:

Lower Paleolithic:

[The Acheulean includes] biface assemblages older than

the Middle Paleolithic, where the Middle Paleolithic includes

all assemblages with or without bifaces that have Mouste-

rian-like flake tools and may be as old as stage 8 in the

oxygen isotope chronology. (Villa 1991, 199)

The majority of specialists recognize in the Acheulean an

assemblage of lithic industries characterized by the presence

of bifaces. (Tuffreau 1996, 7)

In both Africa and Europe the Mousterian/MSA differs

from the preceding Acheulean mainly in the absence of large

bifacial tools (hand axes and cleavers). (Klein 1999, 408)

Middle Paleolithic:

The most basic innovation marking the Middle Paleolithic

remains the emergence of prepared-core technique. (Rol-

land 1988, 179)

Most of these [MP] industries share one or more of a

number of characteristics, including the use of prepared-

core flaking techniques and a limited range of major tool

classes (principally racloirs [i.e., “scrapers”], denticulates/

notches, and bifaces). (Dibble and Rolland 1992)

The prime hallmark of Middle Paleolithic technology

[is] the emergence of more complex and sophisticated pat-

terns of prepared-core flaking, classically illustrated by the

various Levallois and allied techniques. (Mellars 1996, 4)

[The Middle Paleolithic] is a stage anterior to the Upper

Paleolithic where lithic assemblages are characterised by a

high proportion of standardised flake-supports and flake-

tools. These are based on debitage which generally, but not

always, involves the Levallois technique. (Gamble and Roe-

broeks 1999, 5)

By the Middle Paleolithic we mean the period in which,

at least in Europe, the first standard techniques of flake

manufacture were developed (Tuffreau 1979; Bosinski 1982).

(Vega Toscano et al. 1999, 23)

Middle Paleolithic industries, as characterized by the oc-

currence of fully developed Levallois technology, seem to

start about 300,000 years ago. (Roebroeks and Tuffreau

1999, 121)

Well-made Levallois flakes and cores appear in some later

Acheulean assemblages, dating between perhaps 400 and 200

ky ago, but they become common only in Mousterian and

MSA assemblages after 200 ky ago. Not all Mousterian/MSA

people produced Levallois flakes, but many did, and Le-

vallois technology is sometimes regarded as a hallmark of

the Mousterian/MSA. (Klein 1999, 411)

Unfortunately, the continued use of these criteria and the

absence of a serious examination and revision of the definition

of the Lower/Middle Paleolithic periodization have led to a

focus on limited aspects of variation while neglecting other

questions.

For example, the meaning of the presence or absence of

bifaces in industries such as the Acheulean and Clactonian

has been debated ever since bifaces became the defining cri-

terion for the Lower Paleolithic. It has now been demon-

strated conclusively that these two industries are contempo-

raneous (Ashton and McNabb 1996) and that there is no

cultural distinction between them (Ohel 1979; Ashton and

McNabb 1992; Roberts, Gamble, and Bridgland 1995); they

differ only on the basis of the presence or absence of bifaces

(Ashton et al. 1994). In fact, bifaces have even turned up in

Clactonian contexts, providing the final blow to the notion

of the Clactonian as a real tradition (Ashton and McNabb

1996; McNabb 1996; Conway 1995). In other contexts the

meaning of a distinction between early biface- and non-bi-

face-bearing assemblages has been questioned (e.g., Jelinek

1977; Mussi 1995), and many have noted that the only dis-

tinction between certain Acheulean and Mousterian assem-

blages is the presence or absence of bifaces (e.g., Tuffreau

1982; Bosinski in Ronen 1982b; Villa 1991; Bordes 1950a).

Finally, it is becoming increasingly apparent that biface ty-

pology has no consistent temporal patterning (e.g., Bosinski

1995; Tuffreau and Antoine 1995; Roberts, Gamble, and

Bridgland 1995; Milliken 2001; Villa 1991). In other words,

biface morphology and presence/absence seem to vary in-

dependently of any other variable except for raw material (e.g.,

McPherron 2000; Gamble and Marshall 2001). Bifaces are

probably simply a basic component of Lower and Middle

Paleolithic toolkits throughout much of the world and as such

carry a limited amount of cultural and temporal information.

The continued focus on index fossils such as bifaces, which

are so widespread spatially and temporally that they carry

almost no meaning, and their continued use in defining the

Lower and Middle Paleolithic lead to two key questions: (1)

How have archaeologists justified retaining these index fossils

to distinguish the Lower from the Middle Paleolithic? and (2)

Does the diachronic patterning of these artifact types in well-

dated, well-excavated sites support the traditional Lower/Mid-

dle Paleolithic periodization? The first question requires a

historical treatment. The second relies upon detailed infor-

mation from absolutely dated sites in western Europe to trace

the diachronic patterning of each index fossil through time.

Monnier The Lower/Middle Paleolithic Periodization 711

Table 1. De Mortillet’s 1869 Classification of the Paleolithic

Epoch Characteristic Tool-Type Representative Sites

La Madeleine Bone points and har-poons, art

La Madeleine, LesEyzies, Laugerie-Basse, Bruniquel,Massat

Aurignac Split-base bone points,bola stones

Aurignac, Gorge-d’Enfer, Cro-Mag-non, Chatel-Perron,Grotte de la Chaise

Solutre Finely worked bifacialpoints, bola stones

Solutre, Laugerie-Haute, Pont-a-Less

Moustiers Almond-shaped hand-axes, unifacial points

Grotte du Moustiers,Grotte de Pey-de-l’Aze, Grotte del’Ermitage, Chez-Poure, Quaternaryalluvial deposits ofthe Somme Valleyand the Seine Valley

The Identification of Handaxes, LevalloisTechnology, and Flake Tools as IndexFossils

The history of the periodization of the Paleolithic from itsinitial formulation in the 1860s to its essentially modern formin the 1960s reveals that the wide temporal overlap betweenindex fossils was recognized at the time of the very first ar-tifact-based classification of the Paleolithic but no prehistorianever suggested abandoning them. Instead, prehistorians cre-ated increasingly convoluted frameworks to account for theiroverlapping distribution. Further, the decades-old debate onthe significance of biface versus flake industries stems froma dichotomy created in the first classification of the Paleolithicand subsequently reified and amplified through time. Finally,each new classification was the product of a dialectic betweenresearch paradigms and the ever-increasing resolution of thedata.

Gabriel de Mortillet and the Identification of PaleolithicIndex Fossils

The term “Paleolithic” was coined by John Lubbock (1865,2–3) to distinguish the Old Stone Age (“Age of Flaked Stone”)from the New Stone Age (“Age of Polished Stone,” or Ne-olithic). The Paleolithic then began to be subdivided andrefined, initially on the basis of faunal sequences. The Frenchpaleontologist Edouard Lartet, working in the DordogneValley of France, excavated key sites such as Le Moustier, Pechde l’Aze, and La Madeleine and divided the Paleolithic intothree epochs: the epoch of the cave bear (Ursus spelaeus), theepoch of the mammoth (Elephas primigenius), and the epochof the reindeer (Cervus tarandus) (Lartet and Christy1865–75). His classification was initially adopted by anti-quarians and geologists, but they soon favored an artifact-based classification proposed in northern France by Gabrielde Mortillet.

De Mortillet rejected Lartet’s paleontological classificationon the basis of three arguments (Mortillet 1872). First, heargued (1869) that the animal species upon which Lartet’sclassification was based are found throughout the Paleolithic(a limitation Lartet acknowledged) whereas lithic industriesare more variable and provide a stronger foundation for achronology. Second, he argued (1872, 1883) that species abun-dance will be affected by site locality and type (e.g., cave versusopen-air). Finally, he suggested (1872) that the classificationof the Paleolithic, like that of later periods, should be basedupon artifacts, which he felt would provide the best meansof tracing phases of human development. Ironically, the sameproblems he criticized in the faunal classification were laterto plague his own artifact-based classification, as demon-strated below.

Although he eschewed the paleontological approach, deMortillet did borrow two concepts from geology. The first

was the practice of naming periods after the best-known andmost typical locality (Mortillet 1869). Thus, he named theAcheulean after the site of Saint-Acheul in northern France,the Mousterian after the site of Le Moustier in southwesternFrance, and so on. Second, he defined each period on thebasis of a single characteristic tool type or fossile directeur, anotion borrowed from the geological concept of index fossils.De Mortillet’s methodology, therefore, was derived primarilyfrom geology and viewed artifact sequences as analogous tofossil sequences (Sackett 1981, 1991).

As director of the Musee des Antiquites Nationales, deMortillet was asked to classify the prehistoric exhibitions forthe Universal Exposition in Paris in 1867. This led to his firstclassification of the prehistory of France, published in a guideto the exposition “Promenades prehistoriques a l’expositionuniverselle” (Mortillet 1867). In this and a similar classifi-cation published in 1869, he defined the Mousterian as thefirst epoch of the Stone Age and characterized its index fossilsas “almond-shaped handaxes” (also called bifaces) and “uni-facial points” (table 1; Mortillet 1869). In a subsequent clas-sification published in 1873, however, he changed the nameof the first epoch to “Acheulean” and identified its indexfossils as “large almond-shaped stone tools” (bifaces; table 2).He named the following epoch “Mousterian” and character-ized it in terms of “unifacially worked flint points and scrap-ers.” In other words, in the 1873 version he removed handaxesas one of the index fossils of the Mousterian and made themthe sole index fossil of the Acheulean and points, scrapers,and other flake tools of the Mousterian. He did not reject thepresence of handaxes in the Mousterian, which he suggestedpersisted into this epoch from the preceding Acheulean beforegradually becoming extinct. He did, however, insist that re-touched flake tools did not occur in the Acheulean, on which



Ancient DivisionsIndustrialDivisions

Epochs Basedon Industry Principal Sites

Geology andMeteorology Fauna

Polished stone orNeolithic

Polished stone Robenhausian: pol-ished stone axesand barbed flintarrowheads

Robenhausen, Meilen,Mooseedorf, Wangen(lacustrine occupa-tions), Grand Pres-signy (workshop),Chastay (camp oroppidum)

Currentclimate

Domesticated animals;human races verymixed, brachycephalicand dolichocephalic,analogous to moderns

Flaked stone, Archeo-lithic or Paleolithic

Flaked stone withbone tools

Magdalenian: barbedbone arrowheadsand flint blades

La Madeleine, Les Eyzies,Laugerie-Basse, Bruni-quel, Massat, Montre-jeau, Arcy, Schussenried

Postglacial Reindeer, aurochs, andbears very abundant;also mammoths, hye-nas, large felids

Flaked stoneonly

Solutrean: bifaciallyworked laurel-leafflint points

Solutre, Laugerie-Haute,Badegols, Saint-Martind’Excideuil

Cold anddry climate

Brachycephalic and mes-aticephalic man, closeto the modern race(La Laisse, Cro-Mag-non Laugerie-Basse,Solutre)

Mousterian: uni-facially workedflint points andscrapers

Moustiers, Chez Pourre,La Martiniere,L’Ermitage, La MereGrand, Buoux, Neron,Grenelle, Levallois,Clichy, Le Pecq,Montguillan

Glacial; cold andhumid climate

Cave bears, rhinoceros:inferior, dolicocephalicman (Engis, Oimo)

Acheulean: large al-mond-shaped stonetools

Saint-Acheul, Abbeville,Thenne, Sotteville lesRouen, Vaudricourt,San Isidro (Madrid),Tilly, La Ganterie,Sausse and CeilloneValleys

Preglacial temp-erate climate

Hippopotamus, Elephasantiquus; man themost inferior type(Neanderthal, Eguis-heim, La Naulette,Denize)

point he was challenged by several prehistorians during thediscussion following the appearance of his paper (Bourgeoisand Frank in Mortillet 1872).

The debate regarding handaxes as the sole retouched tooltype of the earliest Paleolithic epoch was not resolved, andin 1883, in Le prehistorique, de Mortillet changed the nameof the earliest Paleolithic epoch from “Acheulean” to “Chel-lean” (table 3), identifying as its sole index fossil handaxes.He was apparently convinced by a collector named Ernestd’Acy that the deposits of Saint-Acheul were disturbed (Mor-tillet 1883) because d’Acy’s collections showed mixing in boththe fauna (Elephas antiquus and E. primigenius bones in thesame layers) and the artifacts (Mousterian tools and handaxesin the same deposits). Although the integrity of d’Acy’s col-lections is doubtful today, his arguments at the time weresufficient to persuade de Mortillet to select a “more pure,characteristic locality” to apply to the first Paleolithic epoch:the site of Chelles (Mortillet 1883, 132–33). De Mortillet em-phasized, however, that the name “Acheulean” could still beascribed to what he saw as a transitional phase between theChellean and the Mousterian (1883, 254), denoting industrieswhich contained a combination of more finely worked, per-fected handaxes and some Mousterian tools (see also Mortillet

and Mortillet 1903, pl. 6). Thus, ironically, he rejected Saint-Acheul on the grounds that it was mixed but held that in-dustries containing handaxes and Mousterian flake tools didexist. In the third, posthumous edition of Le prehistorique,the “Acheulean” appeared once again in the table, betweenthe Chellean and the Mousterian, but was little discussed inthe text (Mortillet and Mortillet 1900). Apparently his son,Adrien de Mortillet, had decided to reinstate it as an auton-omous, transitional epoch, as evidenced in other publicationsof the time (e.g., Mortillet and Mortillet 1903; table 4).

De Mortillet identified the Acheulean as a transitional ep-och between the Chellean and the Mousterian for two reasons.First, increasing numbers of collections from Quaternary de-posits found in the quarries and railroad trenches of northernFrance revealed industries containing both handaxes and re-touched flake tools, and he was forced to incorporate thesenew data into his scheme. Secondly, one of the dominantthemes in the social sciences of his time was that culturesevolved gradually and continuously in a unilinear fashion.This view and the notion that technological development wasprogressive were the bases of his classifications. Both themesstemmed from the Enlightenment paradigm, which, thoughdeveloped during the seventeenth and eighteenth centuries in



Times Ages Periods Epochs

Modern Historic Iron Age Merovingian Wabennian, Frankish, Burgonde, GermanicRoman Champdolian, Roman decadence

Lugdunian, RomanProtohistoric Galatian, Etruscan Marnian, Gaulish, 3rd Lacustrian

Hallstattian, Tumulus, 1st IronPrehistoric Bronze Age Bohemian Larnaudian (du marteleur), 2nd Lacustrian

(part)Morgian (du fondeur), 2nd Lacustrian

(part)Stone Age Neolithic, polished stone Robenhausian, 1st Lacustrian, Dolmens

Magdalenian, Caves, Reindeer (most)Solutrean, Reindeer (part), Mammoth

(part)Quaternary Paleolithic, flaked stone Mousterian, Cave Bear

Challean, Acheulean, Mammoth (part),Elephas antiquus

Tertiary Eolithic, fire-split stone Thenaisian

Table 4. De Mortillet and de Mortillet’s 1903 Classification of the Paleolithic

Times Ages Periods Epochs

Modern Historic Iron Age Merovingian Wabenian (Waben)Roman Champdolian (Champdolent)

Lugdunian (Lyon)Protohistoric Galatian Beuvraysian (Mont Beuvray)

Marnian (Marne Department)Hallstattian (Hallstatt)

Bronze Age Tsiganian Larnaudian (Larnaud)Morgian (Morges)

Prehistoric Stone Age Neolithic, Polished Stone Robenhausian (Robenhausen)Tardenoisian (Fere-en-Tardenois)

Quaternary Paleolithic, Flaked Stone Tourassian (La Tourasse)Magdalenian (La Madeleine)Solutrean (Solutre)Mousterian (Le Moustier)Acheulean (Saint-Acheul)Chellean (Chelles)

Tertiary Eolithic, Exploded Stone Puycournian (Puy Courny)Thenaysienne (Thenay)

Western Europe, continued to influence scientific thought wellinto the nineteenth. In fact, some of the major tenets of theEnlightenment, such as that technological as well as culturalprogress was a main feature of human development, meshedparticularly well with the nineteenth century’s emerging the-ory of evolution. According to Trigger, these Enlightenmentideals not only were a prerequisite for the acceptance of Dar-winian evolution but also were subsequently reinforced by thescientific study of evolution, both in biology and in archae-ology, where the latter was seen as demonstrating the realityof progress in prehistoric times (Trigger 1989, 94, 101). Thus,when we place de Mortillet’s classifications of the Paleolithicin this context, it becomes clear that his emphasis on con-tinuous evolution within a framework of technological pro-gress toward perfection was the product of Enlightenment

ideals. By placing the industries which contained a mix ofChellean and Mousterian index fossils between these two ep-ochs and labeling them transitional (Acheulean), he was ableto accommodate data showing that these index fossils over-lapped while adhering to the notion of gradual developmentand continuity from one epoch to the next.

In defining the Mousterian, de Mortillet also emphasizedcontinuity by selecting a type site which “contains Chelleantools which link it to the previous epoch, as well as forms inthe upper levels which link it to the subsequent period” (1883,252). He emphasized technological progress from the Chel-lean to the Mousterian, claiming that handaxes were “big andheavy” in the Chellean and became “lighter and more finelyand elegantly worked” as they approached the Mousterian (p.254). He saw them as slowly disappearing from the Mous-


Table 5. Commont’s 1908 Classification of the Paleo-lithic

Epoch Type Artifacts

Mousterian Levallois technology; numerous re-touched flake tools

Upper Acheulean Lanceolate handaxes; flake toolsAcheulean Limande handaxes; flake toolsChellean Ficron handaxes; other handaxes, uti-

lized flakes, flake toolsPre-Chellean Partly cortical handaxes, utilized flakes

terian, to be replaced by more complex, specialized flake toolssuch as scrapers, points, saws, and blades.

De Mortillet linked the Mousterian and the Chellean notonly through continuity in handaxes but also by suggestingthat the flaking technology of the Mousterian arose in theChellean: “The Chellean instrument is simply the natural rockperfected by flaking on both sides; it is these flakes that gaverise to the Mousterian industry” (1883, 254). Although hementioned the presence of Levallois flakes in some Mouste-rian assemblages, he did not use Levallois technology as anindex fossil of the Mousterian. He firmly believed, however,that the Mousterian was primarily a flake-based industry whilethe Chellean was handaxe-based.

De Mortillet’s classification has been revised and modifiedmany times, but it still forms the basis for the scheme we usetoday. The most lasting impact arose from his choice of hand-axes as the index fossil of the earliest stage of the Paleolithicand his insistence that retouched flake tools were not a sig-nificant component of those early industries. This choice setup the fundamental handaxe/biface–versus-flake-tool dichot-omy that has spurred controversy and debate to this day (e.g.,Warren 1926; Breuil 1932a; Movius 1948, 1949; Ohel 1979;Ashton and McNabb 1992; Schick 1994). What is not fullyunderstood is why he persisted in his assertion that the earliestindustry, the Chellean, did not contain retouched flake tools.Perhaps it was because this was the simplest way of orderingthe Paleolithic record at the time and if he had accepted theco-occurrence of bifacial implements and retouched flaketools in both the Acheulean and the Mousterian, he wouldhave had no other way of distinguishing the two epochs.

Victor Commont: First Modification of de Mortillet’sPaleolithic Classification

Victor Commont, the first prehistorian to modify de Mor-tillet’s classification after his death in 1898, developed amethod for distinguishing the Mousterian from the Acheuleanthat did not involve denying the presence of flake tools inthe earlier industries. In the early 1900s, Commont traveledextensively throughout the Somme Valley in northern France,investigating excavations, road cuts, and quarries for theirartifact contents. He made detailed maps of geological sectionsand carefully noted which types of artifacts were found ineach level in an effort to clarify the river-terrace sequence ofthe valley and the sequence of industrial types within it.Through these rigorous studies, he gradually developed a geo-logical model of the Somme Valley which divided it into fourmajor terraces and identified the types of artifacts found invarious gravel, sand, and loess deposits on each of them.

Since Commont’s field methods were much more rigorousthan de Mortillet’s, he recognized more variation. First of all,in contrast to de Mortillet, he accepted the presence of re-touched flake tools in pre-Mousterian industries such as theChellean. This is probably because he personally observedexcavations and therefore saw what came out of the ground

rather than retrieving artifacts from workers later. He himselfstated that workers were only looking for belles pieces andfrequently rejected flakes (Commont 1906). Secondly, he em-phasized the fact that handaxes occurred across a very longperiod of time, from the pre-Chellean all the way into theMousterian. In other words, he recognized that de Mortillet’sindex fossils—handaxes and flake tools—occurred togetherin all of these industries. Yet, like de Mortillet, he was op-erating under a paradigm of gradual and progressive culturalevolution and therefore needed a means of differentiating thePaleolithic industries while adhering to this paradigm. Thishe did by introducing two changes to de Mortillet’s classifi-cation. First, he posited that handaxe morphology changedgradually through time, becoming increasingly refined andperfected. Thus his first epoch, the Pre-Chellean, was char-acterized by crude, partly cortical handaxes and utilized flakes(table 5). The succeeding epoch, the Chellean, was charac-terized by the ficron: a pointed, elongated bifacial tool witha thick, often cortical butt and sinuous edges. The Acheuleanwas characterized by the limande, an oval, flat, straight-edgedhandaxe, more finely flaked and more refined than the ficron(Commont 1908). Finally, the “Acheuleen superieur,” or Up-per Acheulean, represented the epitome of the Acheulean; thelanceolate handaxes of this period were so finely worked thatflintknappers of his day could not reproduce them (Commont1906). As did de Mortillet, Commont emphasized gradualprogress and continuity, stressing continuous developmentfrom one handaxe index fossil to the next and noting thattransitional forms were common and occurred between allhandaxe types (Commont 1906, 1908, 1913). He also statedthat the flake tools accompanying each industry became morefinely retouched through time (Commont 1906).

The second change that Commont introduced to de Mor-tillet’s classification was in the definition of the Mousterian.As noted above, he was aware that handaxes were also presentin the Mousterian. Thus, the Mousterian contained the samebasic tool types—handaxes and retouched flake tools—as ear-lier industries, and the problem was how it could be distin-guished from them. Commont accomplished this objectiveby introducing a new variable: flake technology. He arguedthat although Acheulean and Mousterian retouched flaketools resembled each other superficially, they could be distin-guished because Mousterian flakes were the product of Le-


Table 6. Peyrony’s 1920 Sequence of Indus-tries at La Ferrassie and Le Moustier

La Ferrassie Le Moustier

Lower AurignacianMousterian Mousterian

[sand, floods]Classical Mousterian:

reindeer, bovidsMTA: reindeer, bovids,

Abri Audi formsAcheulean: hand-

axes, utilizedflakes, no reindeer

Utilized flakes, handaxewaste flakes, noreindeer

vallois technology. The term “Levallois” was used as early asthe 1860s to refer to large, flat flakes discovered in the Le-vallois-Perret suburb of Paris and was briefly mentioned byde Mortillet in Le prehistorique (de Mortillet 1883, 256). Com-mont made Levallois technology the index fossil of the Mous-terian, arguing that certain features of the platform and bulbcould be used to distinguish Mousterian from Acheuleanflakes (Commont 1913, 65). Levallois technology has re-mained the hallmark of the Mousterian to this day.

In sum, Commont’s detailed fieldwork led him to realizethat de Mortillet’s index fossils were too broad, and he refinedthem by introducing two new variables: handaxe morphologyand flake technology. He also added a few epochs but did nototherwise alter the framework of de Mortillet’s classificationor its fundamental emphasis on handaxes and flake tools.Like his predecessor, he was guided by a paradigm of unilinearcultural evolution and progress. More so than de Mortillet,he emphasized continuity between the successive Paleolithicepochs and the gradual development of stone tools towardperfection over time. He stressed that the transition fromAcheulean to Mousterian was slow and gradual, with Levalloistechnology appearing toward the end of the Acheulean andgradually replacing bifaces during the Mousterian (Commont1913, 68). He also argued that Levallois technology itselfevolved throughout the Mousterian, producing massive flakesin the beginning but gradually yielding flakes that weresmaller, thinner, and had a more complex dorsal scar pattern(p. 70). Toward the end of the Mousterian, burins and bladesappeared, foreshadowing the Upper Paleolithic (p. 70). In-terestingly, although Commont believed that Acheulean pop-ulations were descended from Chellean ones, he suggestedthat the Mousterian population was a separate group whichslowly infiltrated the Acheulean population. Perhaps this wasthe best explanation he could devise to account for increas-ingly evident variability within the Mousterian, such as thefact that bifaces, having decreased in frequency throughoutthe Mousterian, sometimes reappeared toward the very endor in Aurignacian deposits (e.g., at Le Moustier, Hauteroche,and Laussel [pp. 129–31]), and discrepancies such as anAcheulean-like industry in the lower Mousterian levels atMontieres (p. 132).

Denis Peyrony: Explaining Variation in Terms of Dual Phyla

Denis Peyrony spent much of his career elucidating the prob-lem of Mousterian variation in southwestern France. ThoughCommont’s contemporary, he lived much longer, and hispublications reveal important differences in his views on cul-tural evolution. Peyrony spent most of his life in the DordogneValley, where he excavated many sites, several with his teacherand collaborator of many years, Louis Capitan, himself a stu-dent of de Mortillet. Among the Middle Paleolithic sites Pey-rony excavated were Le Moustier (Peyrony 1930), La Ferrassie(Peyrony 1934b), Combe-Capelle (Peyrony 1934a, 1943), andLa Micoque (Peyrony 1938). Using data from these excava-

tions, he began to refine the Mousterian sequence, doing forthe Middle Paleolithic essentially what Commont had donefor the Lower Paleolithic in northern France.

Initially, Peyrony adhered to the notion of unilinear culturalevolution, as exemplified in a 1912 publication on the ex-cavations at La Ferrassie (Capitan and Peyrony 1912). Capitanand Peyrony emphasized the continuity of occupation fromthe Acheulean through the Aurignacian and the gradual evo-lution from one industry to the next, with the Mousteriansequence showing increasing perfection of the retouchedtools. This was in keeping with the themes of linear culturalevolution and technological progress through time empha-sized by de Mortillet and Commont.

Peyrony’s views began to change, however, in his solopublications a few years later. In 1920 he published a shortnote in which he suggested that the Mousterian was composedof two distinct but contemporaneous facies: (1) the “ClassicalMousterian,” characterized by Mousterian points and scraperswith invasive retouch and utilized bone flakes, and (2) the“Mousterian of Acheulean Tradition” (MAT), characterizedby numerous handaxes, backed knives, and utilized flakes.This dual-facies model was a radical departure from the uni-linear model of cultural evolution, but Peyrony emphasizedcontinuity by proposing that the MAT was derived from theAcheulean and was analogous to the Upper Acheulean foundby Commont (1908) at Saint-Acheul.

What caused Peyrony to depart from the unilinear culturalevolution model? Along with the prehistorians before him,he seems to have been influenced both by archaeological dataand by overarching scientific paradigms. The archaeologicalrecord of the time was becoming increasingly detailed as ex-cavation and geological methods were refined. Although hismethods were crude by today’s standards, Peyrony generallyexcavated sites more carefully than his predecessors and paidmore attention to the geological formation of the sites. Healso tried correlating sites and began building a regionalchrono-stratigraphic scheme. These endeavors led him to twoconclusions which called unilinear evolution into question.The first was that a Classical Mousterian level (C) at La Fer-rassie was contemporaneous (on the basis of stratigraphy,fauna, and lithics) with the MAT deposits (levels F–H) at LeMoustier (Peyrony 1920, Peyrony 1934 table 6). According


to the existing paradigm, handaxes were an index fossil ofthe Chellean and Acheulean and were supposed to be replacedby flake tools in the Mousterian. The contemporaneity of ahandaxe industry at Le Moustier and a flake tool industry atLa Ferrassie was therefore not reconcilable with a unilinearscheme. The second conclusion was that there were “reversals”of the Acheulean-to-Mousterian progression. In 1930, in areanalysis of the Lower Terrace deposits at Le Moustier, Pey-rony realized that level B was more accurately described asClassical Mousterian than as Acheulean (the term he had usedin 1920). This meant that the MAT industries at Le Moustier(levels F–H) followed a Classical Mousterian industry (levelB). Since Peyrony believed that evolution was progressive, theonly way he could explain such a reversal in the normalcultural sequence was by invoking dual phyla.

The synchronic variation observed by Peyrony was rein-forced by a new paradigm in archaeology. The concept of thearchaeological culture arose in the late nineteenth century andresulted in the culture-historical approach to the study ofprehistory (Trigger 1989, 161). By the 1920s, Trigger argues,cultural evolutionism had largely been replaced by this newapproach, exemplified in the works of V. Gordon Childe (pp.148–50). Childe’s The Dawn of European Civilization, pub-lished in 1925, interpreted European prehistory in terms ofa complex mosaic of cultures and forever changed the faceof European archaeology (p. 168): “The primary aim of ar-chaeologists who adopted this approach was no longer tointerpret the archaeological record as evidence of stages ofcultural development. Instead they sought to identify oftennameless prehistoric peoples by means of archaeological cul-tures and to trace their origin, movements, and interaction”(p. 172). Whether Peyrony was directly influenced by Childe’swork is difficult to determine, but he was clearly a part ofthis new movement. In his writings, unlike those of his pred-ecessors, there are clear references to groups of people—theirorigins, movements, and interactions with each other (e.g.,Peyrony 1930, 43-45).

In sum, Peyrony’s archaeological thinking was affected bythe general shift in the current intellectual paradigm whichmarked early-twentieth-century Western European archae-ology. He still believed in continuous cultural evolution, ascan be seen in his placing the roots of the MAT in the Chellean(Peyrony 1930). However, in proposing multiple facies andin actively attributing them to different populations, he dif-fered greatly from his predecessors (although Commont, forexample, had occasionally referred to populations).

It is notable that Peyrony’s initial observations regardingthe contemporaneity of handaxe- and non-handaxe-based in-dustries did not lead him to question the appropriateness ofthe index fossils, especially handaxes, but instead led him topropose dual facies so that the existing Paleolithic frameworkcould be retained. While he did introduce some new indexfossils, such as backed knives for the MAT and scrapers withinvasive retouch for the Classical Mousterian, the presence orabsence of handaxes remained one of the most critical factors

in assigning an assemblage to one industry rather than an-other and in fact formed the basis for his dual facies.

Henri Breuil: Biface versus Flake Industries

The presence or absence of handaxes or bifaces (as they cameto be more commonly known) was also the basis for HenriBreuil’s classification of the Lower and Middle Paleolithic.Abbe Henri Edouard Prosper Breuil, one of the most influ-ential figures in the history of Paleolithic archaeology, isknown for his synthetic revisions of the entire chronology ofthe Paleolithic, as well as his worldwide travels documentingparietal art. His studies of the Lower and Middle Paleolithicwere published in a series of detailed articles with L. Kozlowskiwhich focused on the stratigraphy of valleys in northernFrance, Belgium, and England (Breuil and Kozlowski 1931,1932a, 1932b, 1934a, 1934b). Essentially, as Peyrony envi-sioned two separate phyla for the Mousterian, Breuil envi-sioned two separate phyla for the whole of the Lower as wellas the Middle Paleolithic. These phyla were based on whathe saw as a natural dichotomy in the industries: some con-tained flake tools and bifaces while others contained flaketools and lacked bifaces (according to Groenen [1994, 199],a similar idea was presented by Hugo Obermaier in his 1925El hombre fosil). This was, to some extent, the basis for Pey-rony’s facies, but Breuil went farther by pushing this dichot-omy back into the Lower Paleolithic and defining more cul-ture groups along each line. Like Peyrony, he linked eachphylum to a distinct population group, but he also linkedeach to certain climates and regions, postulating that the bi-face-making population was centered in southwestern Europewhile the flake-only population was centered in northeasternEurope. As the climate oscillated, he suggested, these popu-lations migrated south during glacials and north during in-terglacials (following the fauna), thereby replacing each otherin various sites. Finally, they fused during the Middle Pale-olithic (Breuil 1932a).

Breuil identified two major industrial traditions in theLower and Middle Paleolithic. Breuil’s flake sequence (1932a;fig. 1) began with the Ipswichian (defined from sites in theEnglish Crags and characterized by small flakes with irregularretouch and variable platforms) and was followed by the Clac-tonian (defined by Breuil 1932b), which gave rise to the Lev-alloisian and the Tayacian, the latter of which culminated inthe Mousterian. His biface sequence began with the Chellean,which he renamed “Abbevillian” when the integrity of Chelleswas questioned on the discovery of an upper Acheulean as-semblage in its lowest layer (Breuil 1932b). The Abbevillian,characterized by massive bifaces and some retouched flakes,was followed by the Acheulean. The sequence culminated withthe Micoquian, which Breuil described as a smaller versionof the Acheulean with very delicate working and many flaketools such as points and scrapers.

Breuil’s scheme was no doubt an attempt, like his prede-cessors’, to reconcile the archaeological data with the legacy


Figure 1. Breuil’s correlation of Paleolithic industries with glacial periods.

of an outdated classification. In addition, Breuil, like Peyrony,was a product of the shift from a cultural evolutionary par-adigm to a culture-historical one. In fact, culture history’semphasis on diffusion, migrations, and so forth, providedsome convenient ways of making his new scheme work. Hisuse of de Mortillet’s index fossils to structure his scheme ledto great difficulties. For example, whereas de Mortillet andCommont had seen bifaces as a perfectly natural componentof the Mousterian, Breuil placed the Mousterian in his flake-industries sequence. However, he was forced to admit thatsome assemblages contained Mousterian flake tools as well asbifaces. Thus, he suggested that the Mousterian and Leval-loisian eventually mixed (interbred) with the Micoquian stageof the biface tradition, producing assemblages which con-tained both bifaces and flake tools (1932a). Another difficultyarose from his separation of Levalloisian and Mousterian in-dustries, which he portrayed as being derived independentlyfrom the Clactonian. In other words, not only did he removeLevallois technology from his definition of the Mousterianbut he made it the defining characteristic of a separate in-dustry. However, he also realized that many assemblages con-tained a combination of Mousterian flake tools and Levalloistechnology, so he reconciled this fact with his scheme by againproposing that the two industries, though separate, affectedeach other through time, resulting in increasing proportions

of Levallois debitage in Mousterian assemblages (1932a). Hisdescription of material cultures’ blending as a result of contactbetween separate populations exemplifies the culture-histor-ical approach. In conclusion, Breuil’s bilineal cultural evo-lutionary scheme had two major difficulties; he had to fit allof the variation into two lines, and the biface/flake dichotomyturned out to be very problematic in that the two variablescrosscut culture types (e.g., some Mousterian assemblages hadbifaces while others did not).

Francois Bordes: Evolution buissonante and the QuantitativeMethod

Breuil’s parallel-phyla scheme was rejected by Francois Bordesin his seminal 1950 article “L’evolution buissonante des in-dustries en Europe Occidentale” (Bordes 1950b). In this paperBordes dismantled the stratigraphical reconstructions ofBreuil, pointed out flaws in his reasoning, and refuted theidea of linear cultural evolution, mono- or diphyletic. Instead,he presented detailed arguments and stratigraphic evidenceto support his complex “branching evolution” scheme (fig.2). The basis for this scheme was the construction of a treeof evolutionary relationships between lithic industries clas-sified on the basis of (1) the presence or absence of bifaces,(2) the percentage of Levallois technology, (3) the percentage


Figure 2. Bordes’s “Evolution of Lower and Middle Paleolithic Industries.”

of faceted platforms, and (4) the Levallois typological index.The resulting classes of industries were then arranged in chro-nological order, with putative evolutionary relationships link-ing the most similar. Bordes retained many of the terms thatBreuil, Peyrony, and other early prehistorians had created toapply to similar types of industries.

Throughout the 1950s, Bordes continued elaborating hismethodology (Bordes and Bourgon 1951; Bordes 1954;1961a). He noted that subdivisions of the Mousterian hadbeen recognized but not adequately defined. In addition, hestated that true index fossils were rare and that assemblagesshould be classified on the basis of many tool types (1961a,2). Therefore, he introduced a methodology for quantifyingassemblage variation based upon relative tool-type frequenciesin his newly created typology. This yielded five main Mous-terian variants which, he argued, represented separate, con-temporaneous cultures that had little interaction with eachother (Bordes and de Sonneville-Bordes 1970). According toBordes, each of these Mousterian cultures had evolvedthrough time, with roots in the Lower Paleolithic, and haddescendants in the Upper Paleolithic (Bordes 1961b; Bordesand de Sonneville-Bordes 1970). For example, he frequentlystated that the MAT-A was derived from the Upper Acheuleanand evolved into the MAT-B and finally the Lower Perigor-

dian. Clearly, then, Bordes, like his predecessors, believed incultural evolution, though he differed in the sense that hesaw it not as mono- or diphyletic but as diversifying throughtime, branching out (buissonante). In fact, this branching no-tion, combined with his view that typological or technologicalcharacteristics were lost or acquired through time (Bordes1950b), made his scheme essentially cladistic in structure.

The source of this difference from his predecessors, as sug-gested by Groenen (1994, 140), was that Bordes was a Dar-winist, trained by Marcellin Boule, whereas his predecessorswere essentially Lamarckian. Thus, rather than envisioningcultures as progressing unilinearly toward perfection, Bordessaw them as diversifying through time, each adapting to theunique constraints imposed by climate, fauna, and the en-vironment. His view of technological progress also differedfrom that of earlier prehistorians. He did not believe thattypes evolved toward perfection. Rather, he believed that toolswere originally undifferentiated and over time became moreand more specialized and standardized (Bordes 1961b, 1970;Bordes, Rigaud, and de Sonneville-Bordes 1972), again anessentially biological evolutionary view. This is the idea ofprogress which still pervades the field today (e.g., Bosinski1996, 167; Moncel and Combier 1992a; Tattersall 1995, 244).

Bordes clearly showed a great deal more sophistication and


detailed knowledge than his predecessors of the variation in-herent in the Middle Paleolithic archaeological record. How-ever, when it came to distinguishing between the Lower andthe Middle Paleolithic, he used a very standard definition,pointing to the presence or absence of bifaces, faceted plat-forms, and Levallois debitage (Bordes 1950b). This oversim-plification is surprising given that he was clearly aware thatLevallois technology could occur in the Lower Paleolithic andbifaces in the Middle Paleolithic. In fact, in the same paper,Bordes described a northern type of Upper Acheulean whichcontained Levallois debitage (Bordes 1950b), and in otherpapers he frequently stated that Levallois technology appearedin the Upper or even the Lower Acheulean (Bordes 1961b,1970). Likewise, he clearly accepted the occurrence of bifacesin some Middle Paleolithic contexts, since one of his Mous-terian variants was Peyrony’s MAT. In sum, while Bordesexplicitly rejected index fossils and proposed that most lithicvariation during the Middle Paleolithic was random and hadno chronological significance, when it came to the LowerPaleolithic and the Lower-to-Middle Paleolithic transition hisoutlook was old-fashioned and involved two index fossils:bifaces and Levallois technology.

The Paleolithic Classification since Bordes’s Time

Around the time that Bordes began defending his Mousteriancultures against American and British critiques (Binford andBinford 1966; Binford 1973; Collins 1970; Freeman 1966; Mel-lars 1965, 1969; Bordes 1969, 1970; Bordes and de Sonneville-Bordes 1970; Bordes, Rigaud, and de Sonneville-Bordes 1972),Grahame Clark proposed a new scheme for classifying lithicindustries based on modes of production (Clark 1969). Thus,Mode 1 industries were defined on the basis of chopper-toolsand flakes and corresponded to the Lower Paleolithic as con-ventionally defined, as did Mode 2 industries, which werecharacterized by bifaces. Mode 3 industries were defined onthe basis of flake tools from prepared cores and correspondedto the Middle Paleolithic. Clark viewed technology as subjectto evolutionary forces such as selection and emphasized pro-gress throughout time as new technologies replaced obsoleteones. He stressed that although not all modes of productionexisted everywhere, their order was inviolable. The strengthof Clark’s scheme was that it conceived of each mode ofproduction as independent and allowed for temporal overlap.In addition, by using the modes to signify technologies ratherthan cultures, his scheme was much more logical when ap-plied throughout the world. Unfortunately, it also reflectedthe notion of progress which was still characteristic of histime, and his insistence that the order of acquisition of tech-nologies was inviolable was clearly inaccurate—some regionsseem to have skipped Mode 2 almost entirely.

Shortly after Bordes’s death in 1981, participants in a con-ference on the Lower/Middle Paleolithic transition organizedby Avraham Ronen in Haifa, Israel, agreed that the boundarycould no longer be equated with the Riss/Wurm boundary

and that the transition was more gradual and continuous thanhad previously been acknowledged. Many advocated pushingthe boundary back into the penultimate glacial (e.g., Bosinski1982; Laville 1982; Tuffreau 1982; Valoch 1982), and somewent farther, arguing that there was no evident cultural breakthat could be used to divide the two (e.g., Bar-Yosef 1982;Jelinek 1982).

Since then, there have been few major revisions of theLower/Middle Paleolithic periodization. In fact, the 1970s and1980s witnessed a shift away from questions of culture historyand diachronic patterning to interpretations of synchronicvariability and cultural explanation driven in large part by theNew Archaeology. In continental Europe, the new techno-logical approach based upon Leroi-Gourhan’s work (1943,1964) began to manifest itself in the late 1980s and becamedominant in the 1990s (e.g., Tixier, Inizan, and Roche 1980;Tixier 1984; Boeda, Geneste, and Meignen 1990). It also cameto be combined with the attribute-analysis approach char-acteristic of Anglo-American archaeology, creating a newmethodology (Tostevin 2003a, 2003b). During this time therehave been few studies on diachronic patterning in the Middleand early Upper Pleistocene. One topic of interest has beenthe timing of the initial colonization of Europe, which haswitnessed a debate between long (Bonifay and Vandermeersch1991) and short (Roebroeks and van Kolfschoten 1995b; Den-nell and Roebroeks 1996) chronologies, with a recent com-promise which accepts incursions of humans into Europeprior to 500,000 years ago (e.g., Parfitt et al. 2005) but nottheir permanent occupation (Dennell 2003). Lately, too, aseries of workshops sponsored by the European Science Foun-dation has focused deliberately on issues within each Paleo-lithic period rather than across transitions (Roebroeks andvan Kolfschoten 1995a; Roebroeks and Gamble 1999; Roe-broeks and Mussi 2000). Most recently, the Middle Paleolithicand Middle Stone Age have become the focus of efforts aimedat identifying the origins of behavioral modernity (McBreartyand Brooks 2000; Henshilwood et a1. 2001; d’Errico 2003).

In summary, while the wide overlap between index fossilswas known from the very beginning, de Mortillet’s Paleolithicclassification, once adopted, assumed an almost tyrannicalhold over systematics and became impossible to remove. Ev-ery prehistorian who grappled with the Paleolithic record hadto modify de Mortillet’s classification in order to accom-modate new data within the index-fossil-based framework.This was done by a variety of means, including renamingepochs on the basis of sites whose typological assemblagesseemed a better fit with the classification, introducing inter-mediate epochs and index fossils to emphasize continuity, andcreating additional facies. Rather than replace the index-fossil-based definitions, these early prehistorians created increas-ingly convoluted schemes in order to retain them.

Given the persistence of these tool types in the definitionsof the Lower and Middle Paleolithic, the questions facing ustoday are (1) What is the actual temporal patterning of thesetool types? and (2) Does it justify a continued division of the


archaeological record into a Lower and a Middle Paleolithicperiod? These questions could not have been answered untilrecently because of the lack of an independent means of as-sessing the chronological patterning of these artifact types.Fortunately, today we have an unprecedented number of ab-solutely dated sites which can be used to evaluate how wellthe actual archaeological record agrees with the index-fossil-based definitions of the Lower and Middle Paleolithic.

Chronological Patterning of the IndexFossils

Assemblage Sampling Strategy

In order to investigate the questions listed above, a databaseof absolutely dated Middle and early Upper Pleistocene as-semblages in Western Europe was created. Strict criteria wereemployed in an effort to produce a database with the greatestpossible integrity. Sites and assemblages were included onlyif they met the following criteria, formulated in order to en-sure typology-free dating, statistically valid sample sizes, in-tegral data, and, finally, feasibility: (1) Assemblages had tocome from sites dated by absolute (radiometric) methods. (2)The minimum assemblage size required was 100 total lithicpieces. (3) Tool counts had to be published or otherwiseaccessible. (4) Assemblages had to come from stratified, re-cently excavated (post–World War II) sites. (5) The geograph-ical limits were Western and Southern Europe.

The first criterion was the most important. Investigatingthe diachronic patterning of tool types required a chrono-logical assessment of assemblages independent of their ty-pological content, and the most reliable and consistent wayof achieving this was to restrict the sample to sites dated byabsolute (radiometric) methods. It is important to note thatthe sample is restricted not to absolutely dated assemblagesbut to absolutely dated sites; if only the former had been used,the sample would have been too small for any meaningfulanalysis. In addition, probably none of the Middle Pleistocenesites in this study have uncontroversial dating. It is rare forthe various methods used to date sites—radiometric, bio-stratigraphic, geochronological, and the like—to coincide toprovide a clear picture of a site’s chronology (for a usefuldescription of some specific issues, see Gamble 1999, 99–110).Undoubtedly the chronological interpretation of some of theassemblages used here will change. Thus, it is important tokeep in mind that in selecting assemblages for this study, carewas taken to ensure that they came from well-dated sites,meaning sites where various dating techniques (absolute andrelative) had been carefully applied in order to obtain themost accurate date possible. In all cases, the final interpre-tation of the excavator based upon all lines of evidence (notjust radiometric dates) was used.

The third criterion was a logistical one. Many sites andassemblages could not be included because their tool countsremained unpublished. In many other instances, tool counts

were incompletely published and crucial elements such as theLevallois index (IL) were missing (e.g., Roebroeks 1988).Where Bordesian counts had not been used (e.g., Singer, Glad-felter, and Wymer 1993; Conway, McNabb, and Ashton 1996;Darlas 1994; Kuhn 1995; Moncel 1999), the published countswere converted to Bordesian counts as carefully as possible,but their true comparability remains open to question. It isunfortunate that many recent Paleolithic site publications,especially those in which the focus of analysis is primarilytechnological, no longer report Bordesian counts (e.g., De-lagnes 1992; Pelegrin 1995; Carbonell et al. 1999). While Bor-desian typology can be criticized in many ways, it is the onlyframework we have for systematically comparing assemblageson the basis of tool counts and other important measures ofvariation such as percentages of Levallois flakes. Its continueduse as a standard for reporting data is therefore vital. Finally,other aspects of lithic variation which should be reportedmore consistently include pebble tools, core types, and flakedefinitions such as minimum size and whether broken orfragmentary flakes are counted separately from complete ones.

Most sites included in the database come from France,Britain, Italy, or Spain. There are several reasons for this focus:first, Western Europe is the region where the index fossilswere initially defined, and thus the patterning used to definethem should be reflected in this region if nowhere else. Sec-ond, a restricted geographical area permits a more thoroughsynthesis of the sites relevant to the research questions. Third,given its long research tradition in this domain, this is oneof the best-documented regions of the world for this timeperiod, and the density of sites here is therefore much greaterthan in most other locations.

Methods

A literature search using these five criteria yielded 89 assem-blages from 26 sites (table 7). Information regarding the con-text, chronology, and nature of the lithic industries of eachassemblage was entered into the database. Then, frequencycalculations of the main index fossils were made using totallithic material as a baseline. It was decided that total lithicmaterial (i.e., all the lithic artifacts from an assemblage exceptfor chips [as defined in each publication]) should be usedrather than other denominators such as Bordes’s real or es-sential count, because the research questions in this studyrequired comparisons between assemblages from time periodswhere amounts of flake versus core tools and retouched versusunretouched tools might be very different. In other words,traditional Bordesian indices, in which frequencies of bifaces,scrapers, and so forth, are calculated as a percentage of theretouched component of flake tools, might not have providedadequate comparability between core-tool-dominated andflake-tool-dominated assemblages. In addition, total lithicmaterial is one of the few quantitative assessments that isalways reported, and it is less subject to interpretation than


Table 7. Sites Included in the Analysis

Site Absolute Dates References

Grotte Vaufrey (France) Levels II–X: to kya74 � 18 246 � 76(Th/U and TL)

Rigaud (1988), Blackwell and Schwarcz (1988),Delpech and Laville (1988)

Orgnac 3 (France) Level 2: 280–300 kya (ESR); Level 6:, to kya (U-se-288 � 82 374 � 165, � 94

ries); Level 8: kya (ESR)309 � 34

Moncel (1999), Moncel and Combier (1992a,1992b), Debard (1988), Debard and Pastre(1988), Falgueres et al. (1988)

Abri Suard (France) Travertine kya (Th/U), Level 53′:101 � 7, –28 kya (Th/U); Level 51:245 � 42

kya (TL)126 � 15

Debenath (1974), Blackwell et al. (1983),Schvoerer et al. (1977), Schwarcz and De-benath (1979), Schwarcz et al. (1983)

Abri Bourgeois-Delaunay (France) Level 7: , kya (Th/U and97 � 6 106 � 10TL); Level 11: kya, (bottom)151 � 15and kya (top) (Th/U),112 � 5 146 �

kya (TL)16

Debenath (1974), Blackwell et al. (1983),Schwarcz and Debenath (1979), Schwarcz etal. (1983)

Biache-Saint-Vaast (France) Level IIA: kya (TL)175 � 13 Tuffreau and Somme (1988)Pech de l’Aze II (France) Levels 3–9: 60–162 kya (ESR) Grun et al. (1991), Bordes (1972)Pech de l’Aze I (France) Level 4: OIS 3 (ESR) Bordes (1972), Soressi et al. (2002)Abri des Canalettes (France) Level 2: 82 kya (TL) Meignen (1993)Combe-Capelle (France) Levels I-1B–I-1E: 48–57 kya (TL) Dibble and Lenoir (1995), Valladas et al. (2003)Fonseigner (France) Levels Ds, Dmi, E: , ,50 � 5 53 � 6 56 �

kya (TL)7Geneste (1985), Valladas et al. (1987)

Grotte du Lazaret (France) Level E travertine: kya (U-series)114 � 61 Darlas (1994), Shen (1985) in Darlas (1994)Cagny-la-Garenne (France) Level I2: kya (ESR/Th/U); allu-448 � 68

vium: kya (ESR [Quartz])443 � 53Tuffreau, Lamotte, and Marcy (1997), Tuffreau

and Antoine (1995), Bahain et al. (2001),Tuffreau (2001a), Antoine (2001), Lamotteand Tuffreau (2001a)

Cagny-l’Epinette (France) Level I1b: to kya318 � 48 231 � 35(ESR/Th/U); Level H: ,253 � 38 260 �

kya (ESR/Th/U)40

Tuffreau, Lamotte, and Marcy (1997), Tuffreauet al. (1995), Bahain et al. (2001), Tuffreau(2001b), Lamotte and Tuffreau (2001b)

Boxgrove (Britain) 175–1350 kya (ESR, OSL, Th/U) Roberts (1986), Roberts et al. (1994), Robertsand Parfitt (1999a, 1999b)

La Cotte de St.-Brelade (Britain) Levels H–5: 190–250 kya Callow and Cornford (1986); Huxtable (1986)Swanscombe (Britain) Lower Loam: 228 kya (TL) Conway (1996), Conway, McNabb, and Ashton

(1996), Bridgland et al. (1985), Stringer andHublin (1999)

Hoxne (Britain) Lower Industry: kya (ESR),319 � 38kya (TL)210 � 20

Singer, Gladfelter, and Wymer (1993), Schwarczand Grun (1993), Bowman (1993), Wymerand Singer (1993), Wymer et al. (1993)

Grotta Guattari (Italy) Levels 1, 4, 5: to kya54 � 4 78 � 10(ESR-LU)

Taschini and Bietti (1979), Schwarcz et al.(1990-91), Kuhn (1995)

Grotta di San Agostino (Italy) Levels 1, 2, 3: to kya43 � 9 54 � 11(ESR-LU)

Tozzi (1970), Schwarcz et al. (1990–91), Kuhn(1995)

Isernia la Pineta (Italy) Level 5: kya (K/Ar)736 � 40 Anconetani et al. (1992), Cremaschi and Per-etto (1988), McPherron and Schmidt (1983),Peretto et al. (1983), Peretto (1991)

Pena Miel (Spain) Level e: kya; Level g 14C:39.9 � 11 45 �

kya–1.2 kya1.4Utrilla et al. (1987)

Gran Dolina TD6 (Spain) kya (paleomagnetism, ESR, U-1 780series)

Carbonell et al. (1999), Falgueres et al. (1999),Pares and Perez-Gonzalez (1999), Carbonelland Rodriguez (1994)

Maastricht-Belvedere(Netherlands)

Unit IV (Sites A–D F–H, K) kya270 � 22(TL); kya (ESR); site K:220 � 40

kya (TL)250 � 22

Roebroeks (1988)

Note: Th/U, Thorium/uranium; TL, thermoluminescence; U-series, uranium-series; ESR, electron spin resonance; OSL, optimally stimulated lu-minescence; LU, linear-uptake; 14C, radiocarbon; K/Ar, potassium/argon; kya, thousand of years BP.

other quantitative measures of lithics such as total numberof flakes.

The Levallois index was used as reported or calculated ac-cording to Bordes’s definition (1950a). The typological indicesused in this analysis are as follows: (1) total lithic material pall flakes, retouched tools, bifaces, pebble tools, and cores; (2)

biface frequency p bifaces/total lithic material; (3) scraperfrequency p scrapers (Bordesian types 9–29)/total lithic ma-terial; (4) flake tool frequency p retouched flake tools (Bor-des’s “essential count” minus choppers and chopping tools)/total lithic material; (5) Levallois index (IL) p Levalloisblanks/Levallois and non-Levallois blanks.


Table 8. Tool-Type Frequencies by Oxygen Isotope Stage

Tool Type

Oxygen Isotope Stage

3 4 5 6 7 8 9 12–14 17�

BifaceN 14 12 7 15 8 11 11 7 4Median 0.00 0.00 0.18 0.17 0.00 0.32 1.70 1.13 0Mean 0.11 0.12 0.17 0.44 0.03 0.38 1.89 1.21 0s.d. 0.21 0.28 0.16 0.85 0.05 0.36 1.36 0.72 0

Flake toolN 14 11 7 15 8 11 11 7 4Median 15.29 12.82 12.68 12.72 7.58 10.71 6.73 5.19 27.49Mean 15.59 13.13 12.55 14.33 7.56 9.59 8.91 4.63 30.26s.d. 10.72 7.56 2.36 7.62 1.05 6.60 6.51 4.69 19.54

ScraperN 14 11 7 15 8 10 11 7 4Median 7.60 7.26 6.14 7.81 3.57 3.93 1.35 0.42 2.65Mean 9.92 8.80 5.87 7.71 3.27 3.83 2.62 0.78 2.35s.d. 8.28 6.71 1.40 3.07 1.28 2.37 3.38 1.03 1.03

LevalloisN 14 12 7 15 8 8 11 7 4Median 0.32 9.71 8.32 19.35 3.09 2.40 0.00 0.24 0Mean 6.67 8.35 9.53 15.40 5.67 4.21 0.04 0.49 0s.d. 10.00 6.23 3.82 11.50 5.52 4.31 0.14 0.64 0

ChopperN 14 12 7 15 8 11 11 7 4Median 0.21 0.09 0.00 0.18 0.01 0.31 0.28 0.16 0.88Mean 0.50 0.97 0.07 0.58 0.03 0.30 0.45 0.31 5.75s.d. 0.91 1.76 0.09 0.71 0.04 0.24 0.59 0.35 10.35

Once the typological indices were calculated, their chro-nological patterning throughout the late Middle and earlyUpper Pleistocene was investigated by plotting average fre-quencies per oxygen isotope stage for all assemblages. Figures3–6 show the median, quartile, and extreme value accordingto oxygen isotope stage for each tool-type frequency as definedabove. Medians and quartiles were chosen because means areoften skewed by outliers in cases of small sample sizes. How-ever, when statistical tests were conducted, these were basedon means and standard deviations (table 8). [All the raw dataused in this study are listed in an appendix in the electronicedition of this issue on the journal’s web page.]

The oxygen isotope stages (OIS) represented are 3–9, cal-culated separately, 12–14 (combined), and 17� (there are noassemblages dated to 15 and 16). While the length of timespanned by oxygen isotope stages varies and therefore mightbe argued to represent inadequately the pace of changethrough time, the greatest difference between time spans—OIS 4, which lasts approximately 10,000 years, and OIS 6,which lasts for about 60,000 years—is still less than one orderof magnitude. The only exceptions are the oldest segments,OIS 12–14 and 17�, in which several stages had to be com-bined because of small sample sizes. However, these combinedstages still showed strong patterning for most of the variablesexamined.

The following assessment examines each of the three mainindex fossil tool types individually in order to determine (1)

its chronological patterning and (2) whether, by itself, it pro-vides support for the Lower/Middle Paleolithic periodization.Subsequently, the three tool types are considered together inorder to determine the existence of any concordances betweenthe three patterns which would justify a division between thetwo periods.

Results

Bifaces

Figure 7, created using the database described above, showsthe temporal patterning of assemblages in which at least onebiface is present and those lacking bifaces. The earliest as-semblages, older than ∼600,000 years (only four are repre-sented here: Gran Dolina TD6 and Isernia La Pineta I-3a, I-3c, and II-3a), do not contain bifaces. From ∼500,000 to∼300,000 years ago, almost all of the assemblages in the sam-ple (15 out of 16; the exception is the Lower Loam atSwanscombe) contain bifaces. Finally, from 300,000 to 35,000years ago, there is a continued occurrence of both assemblageswith bifaces and those lacking bifaces. Boxplots of biface fre-quency in figure 3 (data in table 8) support these presence/absence data: median biface frequencies are 0 in OIS 17�,rise to 1.13 and 1.70 in OIS 12–14 and 9, respectively, andthen decrease in OIS 8 and remain low (but greater than 0)through OIS 3. These patterns seem to confirm the notions


Figure 3. Biface frequency by oxygen isotope stage (lower and upperedges of boxes indicate 25th and 75th percentiles, respectively; bold hor-izontal lines inside boxes indicate the statistical median; asterisks indicateoutliers, circles indicate extreme cases; boxplot widths reflect relativecounts).

Figure 4. Flake tool frequency by oxygen isotope stage (lower and upperedges of boxes indicate 25th and 75th percentiles, respectively; bold hor-izontal lines inside boxes indicate the statistical median; asterisks indicateoutliers, circles indicate extreme cases; boxplot widths reflect relativecounts).

that (1) the earliest European industries do not contain bifaces(e.g., Villa 1991) and (2) biface assemblages are most frequentbetween 500,000 and 300,000 years ago, the time span mostoften equated with the Acheulean.

The data show a statistically significant drop in biface fre-quency from OIS 9 to 8, from a median of 1.70 in OIS 9 to0.32 in OIS 8 (a T-test on mean biface frequencies between

OIS 9 and OIS 8 yields a , , ). TheP p .004 N p 22 d.f p 11.4timing of this event, ca. 300,000 years ago, agrees with mostcurrent definitions of the Lower/Middle Paleolithic boundary.However, there are several problems with using this event tojustify a division into two phases. First, bifaces do not dis-appear after this drop in frequency. They continue to occur,albeit in lower frequencies, throughout the archaeological rec-


Figure 5. Scraper tool frequency by oxygen isotope stage (lower and upperedges of boxes indicate 25th and 75th percentiles, respectively; bold hor-izontal lines inside boxes indicate the statistical median; asterisks indicateoutliers, circles indicate extreme cases; boxplot widths reflect relativecounts).

Figure 6. Levallois index frequency by oxygen isotope stage (lower andupper edges of boxes indicate 25th and 75th percentiles, respectively; boldhorizontal lines inside boxes indicate the statistical median; asterisks in-dicate outliers, circles indicate extreme cases; boxplot widths reflect rel-ative counts).

ord until at least 35,000 years ago. Second, overall bifacefrequencies are extremely low: the highest frequency in a singleassemblage in this sample is 4.93%, occurring in Level I ofCagny l’Epinette. Median biface frequencies for each oxygenisotope stage never rise above 2%. This lack of numericalimportance calls into question the significance of bifaces. Arethey really such an important aspect of Paleolithic variability,or is our continued emphasis on them the result of historical

bias? Further analysis shows that, from a sample of 126 Middleand early Upper Pleistocene assemblages (including some ex-cluded from the database because of a lack of absolute datesbut meeting all other criteria), 49% of them contain no bifaceswhile 51% contain at least one biface. In other words, halfof the assemblages from this time period contain one or morebifaces. The third problem with using the decrease in bifacefrequency between OIS 9 and 8 to support a division into a


Figure 7. Biface presence/absence through time.

Figure 8. Biface percentage frequencies. , ,Mean p 0.43 s.d. p 0.91 n p.126

Lower and a Middle Paleolithic is the number of outliers inassemblages postdating this period. According to the boxplotsin figure 3, there are seven outliers dating to OIS 6 or younger.1

Some of these outliers fall within the frequency range seenin OIS 9 and 12–14. This means that, while biface frequenciesare on average lower after OIS 8, there are situations in which

1. Interestingly, of the seven outlier assemblages with unusually highbiface frequencies which date to OIS 6 or younger, only two are consid-ered MAT: Fonseigner ABC (combined), and Pech I level 4. Levels 8 and9 of Pech II, which contain some of the higher frequencies on the plot,were labeled Acheulean by Bordes. The other outliers are consideredMousterian (Vaufrey Level 1 and Bourgeois-Delaunay Level 9′) or Quina(Pena Miel Level e).

biface frequencies rise again to levels more common beforeOIS 8. These three observations suggest that bifaces, far frombeing a useful chronological marker, are in fact scattered invery low frequencies across the entire European Paleolithiclandscape from 500,000 years ago onwards, occurring inroughly half of the assemblages and achieving peaks of 3–5%frequency in only a tiny fraction of these (fig. 8).

Scrapers and Flake Tools

The presence of retouched flake tools from the earliest as-semblages on is an obvious and indisputable fact today (fig.4), contrary to what de Mortillet argued when he created the


Chellean. The question remains, however, what is the actualtemporal patterning of this tool type? Figure 4 (based on datain table 8) shows that there is a widely fluctuating but gradualincrease in percentage of flake tools through time, from amedian of 5.19 in OIS 12–14 to a median of 15.29 in OIS 3(the extremely high median flake tool frequency in OIS 17�is skewed by three assemblages from Isernia la Pineta andshould probably be treated as an outlier). Interestingly, a me-dian of 12.72 is attained in OIS 6 and remains constantthrough OIS 4.

A similar but slightly less fluctuating chronological patternemerges when median scraper frequencies are plotted throughtime (see fig. 5). This subset of flake tools initially exhibitsvery low frequencies in OIS 17� and especially 12–14 whichgradually rise through time, although they dip fairly stronglyin OIS 7 and 5 (as do flake tool frequencies; the significanceof these dips will be explored below). The more restrictedvariability in scraper tool frequencies probably reflects thenarrower range of types included in the former category. Itwill be interesting to explore the chronological patterning ofother tool types, such as denticulates (which are included inthe flake tool category here), individually. The wide fluctu-ation of these retouched flake tool frequencies provides nosupport for a division of this time period into a Lower anda Middle Paleolithic.

Levallois Technology

A plot of median Levallois index by oxygen isotope stage (fig.6, based on data in table 8) shows that Levallois technologyis almost entirely absent until OIS 8 (traces of Levallois tech-nology have been claimed to occur at Cagny la Garenne andCagny l’Epinette, dated to OIS 12 and 9, respectively). AfterOIS 8, the Levallois index increases, peaking in OIS 6 with amedian of 19.35 before decreasing to a median of 0.32 in OIS3. Interestingly, although median Levallois frequency in OIS3 is very low, the range of values remains large.

Several important observations emerge from these data.First, it is notable that most of the data for OIS 8 come fromone site, Orgnac 3. Most of Orgnac 3’s sequence has beenplaced in OIS 8 (Moncel 1999). However, it is possible thata portion of the sequence may be younger than that, sincethe absolute dates come only from the bottom of the sequence,and the standard deviations for the Th/U dates are very large(Falgueres et al. 1988). The dating is key to establishing Org-nac 3 as the earliest locale where Levallois technology wasused on a regular basis, and full publication of the dates isawaited. Evidence for the next-earliest significant appearanceof Levallois technology occurs only toward the end of OIS 7,at Biache-Saint-Vaast II and Maastricht-Belvedere (Levalloisfrequencies from Maastricht-Belvedere are not published,however). The second observation of note is that Levalloistechnology becomes full-blown only in OIS 6, where the me-dian Levallois index is 19.35. In OIS 7 and 8 the medians areonly 3.09 and 2.40, respectively. This means that only 2–4%

of all blanks are typologically Levallois during OIS 7 and 8,a percentage which many would interpret as too low to bethe product of Levallois technology. The final observation isthat, while some assemblages (such as levels V–VIII of theGrotte Vaufrey) achieve extremely high values for the Levalloisindex, many assemblages between 300,000 and 35,000 yearsago do not contain Levallois flakes at all. This is especiallytrue in OIS 3, where some assemblages have a Levallois indexas high as 20 (e.g., Fonseigner Levels D superieur and D milieu)but most have an index of less than 1. This reduction in theLevallois index at the end of the Middle Paleolithic has pre-viously been noted (Rolland 1988; Mellars 1969; 1996,185–86), and it would be interesting to determine what exactlyis replacing it and why this technological shift occurred. Insum, Levallois technology is absent before OIS 8, becomesfull-blown in OIS 6, and then disappears from many assem-blages by OIS 3 although it remains quite high in a few sites.

The most widely used current boundary for the Lower toMiddle Paleolithic division is 300,000 years ago, or the OIS9/8 boundary. The change in Levallois flake frequencies acrossthese two periods is, in fact, statistically significant (P p

, , ). However, as noted above, the evi-.029 N p 19 d.f p 7.0dence for Levallois technology during OIS 8 is limited. A moreconvincing transition is the dramatic rise in the Levallois in-dex from OIS 7 to 6, which is also statistically significant (T-tests yield a , , ).P p .012 N p 23 d.f. p 20.9

Concordance of Chronological Patterning across the ThreeArtifact Types

When the chronological patterning of all three of the tooltypes traditionally used as index fossils for the Lower and theMiddle Paleolithic is examined, what kind of support is gen-erated for a Lower/Middle Paleolithic division? Two impor-tant chronological events have been identified in the previousdescriptions of the patterning of individual tool types. Thefirst is the drop in median biface frequency from OIS 9 to 8.The second is the rise in median Levallois frequency fromOIS 7 to 6. To what extent are these events supported by thepatterning of the other tool types? There are no statisticallysignificant changes in frequencies of retouched flake tools andscrapers across the OIS 9/8 boundary, but there is a statisticallysignificant increase in Levallois flake frequencies. In fact, Le-vallois index and biface frequency are inversely correlated( , , Pearson’s ). This intriguingP ! .005 N p 115 R p �.344relationship raises further questions. Do Levallois technologyand bifaces form end points of a single axis of variation? DoesLevallois technology originate in biface technology and even-tually replace it, as some have suggested (e.g., Leroi-Gourhan1966 in Rolland 1988; Tuffreau, Lamotte, and Marcy 1997)?Further work will be needed to confirm and elucidate thisinverse relationship.

The second important chronological event identified aboveis the increase in Levallois flake frequency from OIS 7 to OIS6. Interestingly, biface frequencies also rise across this tran-


Table 9. Mean Tool Frequencies by Cold (3, 4, 6, 8, 12) orTemperate (5, 7, 9) stage

Climate N Mean S.D. Std. Error Mean

% bifaces Warm 26 .8565 1.25303 .24574Cold 56 .3046 .53831 .07193

IL Warm 26 4.3292 5.32049 1.04343Cold 53 8.7081 9.65978 1.32687

% scrapers Warm 26 3.6968 2.71919 .53328Cold 54 7.3117 6.01765 .81890

% flake tools Warm 26 9.4750 4.74862 .93128Cold 55 13.0018 8.27035 1.11517

% choppers Warm 26 .2162 .42705 .08375Cold 56 .5867 1.00265 .13399

sition, though not statistically significantly. Finally, there is astatistically significant rise in scraper frequencies from OIS 7to 6 (T-tests yield a , , ). In fact,P p .000 N p 23 d.f. p 20.3OIS 6 contains the highest median Levallois and scraper fre-quencies in the sample. Do these patterns signal a majortransition, or was OIS 6 simply a period that witnessed un-usually high frequencies of these tool types for some unknown(possibly climate-related) reason? In sum, equally significanttypological changes occur across both OIS 9/8 and OIS 7/6,although the earlier event better agrees with the Lower/MiddlePaleolithic periodization. What other trends occur across theMiddle and early Upper Pleistocene among these tool typesin this region?

One pattern that emerges is that many tool-type frequenciesare higher during cold periods (OIS 3, 4, 6, 8, 12) than duringtemperate periods (OIS 5, 7, 9; see table 9 and fig. 9; the patternis reversed for bifaces). T-tests comparing mean tool-type fre-quencies between cold and temperate periods are significantfor bifaces ( ), Levallois index ( ), scrapersP p .039 P p .011( ), flake tools ( ), and “chopper/chopping-P p .000 P p .018tools” ( ). The pattern is especially clear for flake tool,P p .022scraper, and chopper frequencies (see figs. 4, 5, and 10). Themost logical explanation for this trend is the scraper reductionmodel (Dibble 1987, 1995). Specifically, measures of scraperreduction are greater when access to raw material is diminished(Dibble 1991). During cold periods, access to raw material waspresumably diminished by snow or ice cover. Therefore, scrap-ers were resharpened more, and available blanks were trans-formed into scrapers at a greater rate than during warm periods(Rolland 1981; Rolland and Dibble 1990). This pattern wasobserved by Laville (1973), who noted that Denticulate Mous-terian industries characterize temperate periods whereas Quinaindustries are most common in cold periods. Diminished accessto raw material during cold periods may also be the best ex-planation for higher chopper frequencies, although it is alsopossible that certain functions existed during cold periods (e.g.,disarticulating large mammals or frozen carcasses) that requiredmore heavy-duty tools. Mauran, a specialized open-air, bovid-hunting site dated to OIS 3 whose lithics are dominated bydenticulates and chopper/chopping-tools (Jaubert et al. 1990;Jaubert 1993), provides evidence for such a scenario.

The high frequencies of Levallois flakes, scrapers, and bi-faces during OIS 6 may also partly reflect the climatic rigorsof that time period. Many OIS 6 assemblages, such as thosefrom Abri Suard, Grotte du Lazaret, and Pech II, used to belabeled “Acheulean.” This was probably because of the pres-ence of bifaces and the often rough appearance of the re-touched flake tool component. For example, Bordes describedthe flakes and tools from Layer 9 of Pech II as “often batteredby cryoturbation and difficult to characterize. . . . Side scraperswere not very numerous ( ) and usually rather poorSI p 20. . . . There were 10 handaxes ( ), rather crude” (Bor-HI p 9.1des 1972, 56). The extremely cold climate of OIS 6 may haveincreased intensity of occupation and, hence, use of raw ma-

terials and may also have reduced raw-material quality (byresulting in frost fractures) or availability.

Another interesting diachronic pattern evident in these datais the gradual increase in scraper frequencies through time(see fig. 5). The preferred explanation here also involves in-tensity of raw-material utilization. It is quite possible thatover this immensely long span of time, raw materials in thevicinity of sites were exhausted because of intense, long-termoccupation of sites (perhaps related to a shift from open-airto sheltered sites), increasing population densities, or a com-bination of these factors. Decreased raw-material availabilitywould have resulted in greater frequencies of retouched tools,since the supply of blanks would have diminished (Dibble1991; Dibble and Rolland 1992). Alternatively, but more dif-ficult to prove, the trend of increased scraper frequenciesthrough time could reflect technological or cultural factorssuch as new uses or changing cultural norms surrounding theproduction, use, and curation of stone tools.

Finally, there are diachronic trends in the patterning of sitetypes and geographical distribution. Most sites which predateOIS 8 are open-air, whereas most of the younger ones aresheltered. In addition, the earliest sites (earlier than 650,000years ago) are Mediterranean, those dated to 550,000 to300,000 years ago come from northern France and Britain,and those younger than 300,000 years ago come mostly fromsouthwestern France, Spain, and Italy. This patterning maytell us something about the nature of the Lower and MiddlePleistocene occupation of Europe, but it may also bias theresults of this study in unintended ways (although no cor-relations were found between any of the three artifact typesand site type [open versus sheltered]). In summary, there area number of temporal trends which crosscut the broad Lower/Middle Paleolithic scheme and deserve further investigation.These trends represent only the tip of the iceberg; there aredoubtless many other trends reflecting different aspects oflithic variability besides those included here which deservefurther study.

Conclusions

Two questions have been asked in this paper. The first is how


Figure 9. Tool frequency by temperate versus cold periods (diagonalstripes, bifaces; black, choppers; white, scrapers; horizontal stripes, flaketools; vertical stripes, Levallois index).

Figure 10. Chopper frequency by oxygen isotope stage (lower and upperedges of boxes indicate 25th and 75th percentiles, respectively; bold hor-izontal lines inside boxes indicate the statistical median; asterisks indicateoutliers, circles indicate extreme cases; boxplot widths reflect relativecounts).

archaeologists have justified retaining the original index-fossil-based definitions of the Lower and the Middle Paleolithic de-spite increasing evidence of their widespread overlapping tem-poral distributions. We have seen that early prehistorians addedindustries and facies to accommodate this overlap, creatingincreasingly convoluted frameworks in an attempt to reconcilethe data with de Mortillet’s classification. Bordes was the firstto reject index fossils when he redefined the Mousterian facies

on the basis of type frequencies, but even he retained them tocontrast the Lower and the Middle Paleolithic.

The second question is whether the diachronic patterningof these artifact types supports the traditional Lower/MiddlePaleolithic division. This study shows that between OIS 9 andOIS 8, biface frequencies drop while Levallois flake frequenciesrise; the timing of this event agrees with the current Lower/Middle Paleolithic periodization. However, and equally sig-


nificant, between OIS 7 and OIS 6 Levallois, scraper, andbiface frequencies rise. Both events are important; how shallwe privilege one over the other to divide the period into twophases? Conversely, are they an adequate basis for a period-ization? Other diachronic trends have also been identified,such as the drop in Levallois frequency from OIS 6 to OIS 3and the higher frequencies of most types during cold periods.These observations suggest that much temporal variation dur-ing the Middle and early Upper Pleistocene remains to bediscovered.

In sum, it is clear from the history of the Lower/MiddlePaleolithic periodization that we have known from the be-ginning that these three artifact types do not pattern into atwo-phase periodization. The analysis of current data con-firms this beyond a doubt. Bifaces are useless as chronologicalor cultural markers; they are spread across the European land-scape in 51% of assemblages in very low frequencies from500,000 years ago to the beginning of the Upper Paleolithic.Scrapers and other retouched flake tools are not useful indistinguishing the Lower from the Middle Paleolithic becausethey occur in the earliest sites and their frequencies graduallyrise through time, with no clear breaks. Although they arefrequently claimed to be more “standardized” in later periods(Tuffreau 1982; Valoch 1982; Roe 1982; Moncel and Combier1992a; Hayden 1993; Gamble and Roebroeks 1999), evidencefor this has so far been lacking (Monnier 2006). Finally, theappearance of Levallois technology could be used to distin-guish the Lower from the Middle Paleolithic, but then thisboundary would have to be moved to the beginning of OIS6 or somewhere within OIS 7, when Levallois technologybecomes numerically significant.

We must now define a behaviorally meaningful periodi-zation of the archaeological record of the Pleistocene occu-pation of Europe. The first step in this process has beenaccomplished here by demonstrating that the current systemis flawed; the index-fossil-based definitions of the Lower andMiddle Paleolithic should be abandoned. It is time for a com-prehensive revision of the Lower/Middle Paleolithic period-ization based upon a synthesis of multiple aspects of the ar-chaeological record, including climate, subsistence, landscapeuse, mobility and exchange, symbol use, cognition, andbiological evolution, in order to determine whether we shouldmaintain a two-phase system and, if so, how it should bedefined.

Acknowledgments

The core of the research presented here was originally carriedout as part of a Ph.D. dissertation at the University of Penn-sylvania (Monnier 2000), and I thank the members of mycommittee, Harold Dibble, Philip Chase, and Robert Preucel,for their guidance during that process. In addition, I ac-knowledge the financial assistance during my graduate studiesprovided by a William Penn Fellowship and a Traveling Fel-lowship from the French Institute for Culture and Technology

at the University of Pennsylvania. Much of the historical re-search presented here was carried out at the Museum Libraryat the University of Pennsylvania and especially at the TozzerLibrary at Harvard University. The excellent collectionshoused by both these institutions and the kind assistance ofthe librarians made this study possible. Finally, I thank HaroldDibble, Sally Kohlstedt, Greg Laden, Shannon McPherron,Martha Tappen, Gilbert Tostevin, and four anonymous re-viewers for providing useful comments and suggestions forimproving the paper and Nick Ashton for supplying sometool counts from Swanscombe.

Comments

Michael S. BissonDepartment of Anthropology, McGill University, 855Sherbrooke West, Montreal, Quebec, Canada H3A 2T7([email protected]). 22 IV 06

In arguing for a reevaluation of the concepts of the Lowerand Middle Paleolithic, Monnier joins a growing list of schol-ars who are questioning the analytical paradigms that paleo-anthropologists have inherited from nineteenth-century bi-ology, paleontology, and geology. As someone who hasrecently advocated a replacement of the Bordes Middle Pa-leolithic artifact typology with a less ambiguous attribute-based system of artifact description (Bisson 2000), I am en-tirely sympathetic with her goals. As Monnier notes, theperiodization of the Paleolithic coalesced in the 1960s andhas not changed appreciably since, and the creation of theseperiods during the culture-historical phase of archaeology hasimposed an arbitrary boundary that implies a “transition”between the two whereas the course of technological evolutionwas undoubtedly more complex. This is a timely reminderfor professionals, and the historical analysis will be a usefulsynthesis for students.

The diachronic quantitative analysis of the major “indexfossil” types is also a useful contribution, but it is subject tosome important flaws in the nature of the samples chosenand in the assumptions that potentially undermine one of itskey conclusions. These flaws all stem from a failure to dis-tinguish between typology and technology and to take intoaccount the inherent differences in the ratios of “finishedproducts” to unmodified flakes produced by different tool-making strategies. These problems do not undermine theoverall conclusion that differentiating between the Lower andMiddle Paleolithic on the basis of the presence or absence ofartifact “index fossils” is a bad idea, but they do obscure somepatterns that are evident in these data.

Although the criteria applied to assemblages for inclusionin the quantitative analysis are generally good, assemblage sizeis potentially an issue. The appendix in the electronic version


omits the total number of specimens in each assemblage,making assessment of sample-size effects impossible. Moreimportant is the choice to base these comparisons on thepercentage of retouched (or technological, in the case of theLevallois index) pieces in the total of lithic artifacts, includingunmodified pieces. Monnier admits that there may be inter-classifier variability in what is listed in artifact inventories.Did each source list every tiny retouch flake, or did it fail tocount pieces below an arbitrary size requirement (2 cm max-imum dimension)? This is possibly significant because dif-ferent technologies and different end products have the po-tential of producing significantly different numbers ofunmodified flakes or flake fragments per “formal” toolcategory.

My impression as a flint knapper is that creating an ovalhandaxe will produce more debris over 2 cm than producinga retouched uniface such as a scraper, point, or denticulate.As a quick check, I inspected five Acheulean biface forms(oval, subtriangular, lanceolate, Micoquian, and ficron) fromFrance, ranging in length from 9 to 18 cm. Counts of scarsover 2 cm averaged ca. 20 per specimen, whereas an averageof scars over 2 cm on five large French Middle Paleolithicscrapers was only 3.9. Differences such as these could accountfor the fact that bifaces always make up a small proportionof the total artifact inventory compared with unifaces. Thusthe statement that the “lack of numeric importance calls intoquestion the significance of bifaces” needs rethinking. Mon-nier seems to be trapped by typological categories rather thanthinking of the complex interrelationship between typologyand technology in assemblage composition. It is clearly truethat when presence/absence is used, bifaces are indeed “uselessas chronological or cultural markers,” but presence/absenceis an extremely crude analytical tool. A more appropriatecomparison would be the ratio of bifaces to other retouchedpieces, a comparison that could be made independent of theconstraints inherent in the specific flake-tool categories ofBordes.

This criticism notwithstanding, I think that this paper is auseful step in revising how we conceive of cultural processesin early prehistory and how they are described. Categorizationis inherent in human thought and language. It is not sur-prising that categories defined more than a century ago onincomplete data need to be revised. Abandoning the terms“Lower” and “Middle” Paleolithic probably won’t do muchgood, but the overlapping typological contents of the Lowerand Middle Paleolithic illustrated here show that typologicallybased definitions will never be adequate. If we are to use anyarchaeological criteria, they should probably be technological,but given the changing definition of techniques that we haveseen over the last 30 years even this may be problematic.Perhaps we should arbitrarily rely on chronology. Definingthe “Lower Paleolithic” as anything dated to OIS 9 and older,with perhaps an “Intermediate” period of OIS 8 and 7, andthen the “Middle Paleolithic” proper as beginning at OIS 6would free us from the typological straitjacket and allow us

to concentrate on the important questions of explaining theconsiderable variability in the Paleolithic archaeological rec-ord in terms of charging behavior and adaptation.

Richard G. KleinProgram in Human Biology, Bldg. 80, Stanford University,Stanford, CA 94305, U.S.A. ([email protected]).25 IV 06

I agree with Monnier that neither flake tool typology norLevallois technology separates the Acheulean from the Mous-terian, but I believe that biface forms commonly do. TheMousterian of Acheulean Tradition is named for its bifaces,but these tend to be triangular and heart-shaped handaxesthat can usually be distinguished from the often larger, morevariably shaped handaxes that characterize Acheulean assem-blages (Bordes 1961a). Its assemblages also usually lack otherkinds of Acheulean bifaces, particularly cleavers. A problemarises mainly because Acheulean people need not have leftbifaces at every site and some of their contemporaries neverproduced them. In regions that the Acheulean Tradition neverpenetrated, including east-central and eastern Europe (Bos-inski 1995), it is harder to separate the Lower and MiddlePaleolithic on typology alone.

Although Monnier focuses on France and neighboringcountries, her argument raises the broader question whetherwe can recognize culture-stratigraphic units anywhere beforethe Upper Paleolithic. At least in Africa, western Asia, andEurope, I think we can, and the key issues are how abruptlythese units succeeded one another and what relation they hadto human biological evolution. As a working hypothesis, Ioffer the following outline:

The oldest stone artifact tradition, the Oldowan, appearedsuddenly about 2.6 million years ago (Semaw et al. 1997),and it was largely if not exclusively confined to Africa. Ol-dowan assemblages consisted mainly of sharp-edged flakesand the cores from which they came. Homo habilis (or oneof the variants into which it may be split) is usually assumedto have made Oldowan tools. The Acheulean Tradition wasestablished in eastern Africa by 1.65 million years ago (Quadeet al. 2004; Roche et al. 2003), and it was distinguished fromthe Oldowan by the presence of bifaces and of large flakesthat were often made into bifaces. H. ergaster (or African H.erectus) probably invented the Acheulean, and future researchmay show that H. ergaster and the Acheulean emergedabruptly together 1.8–1.7 million years ago. It remains pos-sible, however, that H. ergaster appeared first and that theOldowan continued alongside the Acheulean for hundreds ofthousands of years.

Numeric ages and biostratigraphy show that the AfricanAcheulean may be divided into early and late stages (Clarkand Schick 2000; Isaac 1975; Schick and Clark 2000; Wynn2002). Late Acheulean bifaces were often thinner, more ex-tensively trimmed, and more symmetrical than their earlyAcheulean predecessors, and they were accompanied by a


wider range of well-made flake tools. Late Acheulean assem-blages also varied more through time and space, and it waslate Acheulean people who devised Levallois technology. Thetransition from early to late occurred sometime between 1million and 500,000 years ago, and future research may fix itnear 600,000 years ago, perhaps at the same time as the in-crease in encephalization that ushered in H. heidelbergensis(or “archaic” H. sapiens). H. heidelbergensis took late Acheu-lean artifacts to Europe 600,000–500,000 years ago, and lateAcheulean technology may partly explain how H. heidelbergensisand its descendants became the first permanent Europeans.

The late Acheulean differed from the early more than itdid from the industries that succeeded it—the Mousterian(or Middle Paleolithic) in Europe and western Asia and theMiddle Stone Age (MSA) in Africa. The European Acheuleanis all “late,” and Monnier could cite this to buttress her ar-gument for substantial similarity to the Mousterian.

The Middle Paleolithic/MSA may have supplanted the lateAcheulean at different times in different places, but the avail-able dates suggest that both were in place everywhere by250,000–200,000 years ago (Bar-Yosef 1995; Conard and Fi-scher 2000; Tryon and McBrearty 2002). Assemblage variationis greater within the Middle Paleolithic and the MSA thanbetween them, and their separation depends mainly on schol-arly tradition and geographic distance. They both differedfrom the late Acheulean primarily in the absence of largeAcheulean bifaces, and they shared an emphasis on refinedflake tools and often but not always on Levallois technology.If all we had were the artifacts, we might infer that MiddlePaleolithic and MSA people were physically the same. How-ever, by 160,000 years ago MSA people in Africa were ana-tomically modern or near modern and their European MiddlePaleolithic contemporaries were Neanderthals. About 50,000years ago, the Later Stone Age (LSA) descendants of MSApeople expanded rapidly from Africa to swamp or replace theNeanderthals and other nonmodern Eurasians. LSA peoplewere fully modern in both anatomy and behavior, but spe-cialists disagree about whether the LSA appeared abruptlyabout 50,000 years ago (Klein and Edgar 2002) or developedgradually over a long period within the MSA (McBrearty andBrooks 2000).

John McNabb and Hannah FluckCentre for the Archaeology of Human Origins, Departmentof Archaeology, Avenue Campus, University ofSouthampton, Southampton SO17 1BF, UK ([email protected]). 25 IV 06

Monnier’s use of the Bordes (1961a) system concerns us,though her defence that it allows site comparison over largedistances will strike a chord with most researchers. One of us(JM) has been analysing British Lower Palaeolithic assem-blages for nearly 25 years and has repeatedly found that thesystem’s complexity fails to describe the conservative nature

of these assemblages. Moreover, it does not address multipletools, flaked flakes (only some of which are typologically Clac-tonian notches) (Ashton, Dean, and McNabb 1991), orwedges and nondiagnostic forms, which should not be con-veniently buried in the miscellaneous categories. We recognizethe problems of comparability, but for the British data thesystem is irrelevant.

Similarly, Monnier’s inclusion of debitage in calculationsof indices and percentages, even for the sake of comparability,is inappropriate. Frequencies of flakes (broken or otherwise)reflect site function, hominin transport behaviour, excavationhistory, resharpening and reuse, and many other factors. Withintersite debitage frequencies being so variable, it would bepossible artificially to inflate or depress the true statisticalimportance of significant tool types.

This highlights another failing of the Bordes system. Itsapplication requires an assemblage to include at least 50 toolsand preferably more than 100. In Britain it is rare to haveeven 50, and most are flaked flakes (Ashton, Dean, andMcNabb 1991). It would be better to have calculated theindices on the basis of all the shaped and modified pieces.

Monnier’s list of criteria for site inclusion is laudable, butemployed in the cause of comparability it has led to someserious errors. A combined mammal (Schreve 2001) and geo-logical (Bridgland 1994) signal places Swanscombe in OIS 11,not 9. Radiometric age determinations have consistently un-derestimated the age of the Barnfield Pit sequence. Hoxne isalso an OIS 11 site (Schreve 2000, contra Bowen et al. 1989),and so are the Clacton Golf Course site (Singer et al. 1973)and the West Cliff sections at Clacton (Bridgland et al. 1999).We could add Barnham (Ashton, Lewis, and Parfitt 1998),High Lodge in OIS 13 (Ashton, Cook, and Rose 1992), Pont-newydd in OIS 7 and 6 (Aldhouse-Green 1998, 2001), andeven the recently published sites at Foxhall Road (White andPlunkett 2004) and Elveden (Ashton et al. 2005), both OIS11. Omitting sites such as these because they do not matchthe check-list for consistency is a reflection of the weaknessof some of the selection criteria rather than the sites.

From the above it might be supposed that we dispute Mon-nier’s findings. We do not. We believe her conclusions arevalid; indeed, she provides empirical data supporting a patternthat has often appeared in the literature only in anecdotalform—that bifaces decrease in frequency over time as scrap-ers, other flake tools, and Levallois increase. In the Britishsequence the patterning in Levallois first appears at the endof OIS 9 at sites like Purfleet (Schreve et al. 2002), whereboth Levallois (White in Schreve et al. 2002) and the simpleprepared-core variant (White and Ashton 2003) are present.It is seen again in late OIS 8 and OIS 7 contexts (e.g., WestThurrock, Bakers Hole, and Crayford [see Bridgland 1994 forsummaries]). The key point here is the presence of a persistentprepared-core-technology/Levallois signal from the north-western limit of the hominin world prior to the OIS 6 take-off seen in Monnier’s data. It does not invalidate Monnier’sinterpretation, but it must be taken into account.


Figure 1. A reworking of Monnier’s data. (Hoxne and Swanscombe havebeen placed in OIS 11.) Solid line, percentage scrapers; dotted line, per-centage bifaces; broken line, percentage flake tools. The apparent crashin OIS 10 reflects the absence of data from that period.

The decrease in bifaces and increase in flake tool andscraper frequencies from OIS 6 onwards seems a reasonableresult given what we know about these assemblages. The linkbetween increase in flake tool frequency and the necessity ofutilizing more blanks from pre-existing debitage on cavefloors and resharpening as raw material becomes scarcer dur-ing glaciations is not in itself unreasonable. We are concerned,however, by Monnier’s use of the OIS curve. OIS 3 and OISsub-stages 5.1–5.4 are not interglacials; to compare data fromthem with those from OIS 5.5, 7, 9, 11, and 13 is whollyinappropriate (see our fig. 1).

We agree with Monnier that the Lower/Middle Palaeolithicboundary is an anachronism that is no longer helpful, butwe would caution against downgrading the significance of thetransition to Levallois/prepared-core-technology-dominatedassemblages, which represents a significant component of amultidisciplinary definition of assemblage character thattracks genuine changes in behaviour.

Marie-Helene MoncelInstitut de Paleontologie Humaine, 1 rue Rene Panhard,75013 Paris, France ([email protected]). 20 IV 06

The transition between Lower and Middle Palaeolithic raisesquestions about what criteria best characterize them whenonly preserved lithic assemblages are available to our under-standing. What indicators are relevant to describing the be-ginning of the technical world linked with the Pre-Neander-

thals? Should we focus on the shift to flake production as themajor knapping activity and the diversification of core re-duction methods or on bifaces, which may be missing fromsome Lower Palaeolithic assemblages? The earliest bifaces ofWestern Europe date from more than 0.6 million years ago(Despriee et al. 2005). Though tool kits are common, theassemblages display wide variation that certainly reflects avariety of environments and activities. For example, raw ma-terials distinguish the great northern plains, with large tools,bifaces, and hand-axes/cleavers shaped on flint nodules, fromsouthern Europe, with large flakes. Central European assem-blages follow another trend, resulting in traditions such asthe Micoquian at about OIS 6 at least.

Southeastern French sites provide significant examples ofwhat happened during the second part of the Middle Pleis-tocene. Monnier suggests two phases of transition dated tothe OIS 9/8 and OIS 7/6. At the site of Orgnac 3, the Levalloismethod was the main one in use from OIS 8 onwards. Bifacesare rare throughout the sequence and all but disappear in thelatest occupations. The Mousterian of Acheulean Traditiondoes not occur in this part of France. The Levallois methodappears at Baume Bonne at the end of OIS 8, but its maindevelopment took place in the second half of OIS 6 (Gagn-epain and 2005). In all the layers of these two sites, the mainactivity was flake production. The development of long, com-plex, and varied core reduction sequences is suggested to bethe best indicator of changes in the relationship between hu-mans and raw materials, and these changes are considered


the best criteria for describing a gradual transition rooted inthe oldest flaking traditions (e.g., at Orce, dated to 1.2 millionyears). However, at Payre, dated to OIS 7 and 5, and in thelower levels of Abri Moula (OIS 5) the processing system isdiscoid. The increasingly numerous sites around the RhoneValley during OIS 4 and the beginning of OIS 3 provideevidence for the simultaneous use of Levallois, discoid, andlaminar methods (the laminar method is first attested duringOIS 4 in southern Europe). The Levallois method was stillbeing employed during the late Middle Palaeolithic, for ex-ample, at Abri Moula and Abri du Maras (Moncel 2003).Flake tool frequency does not seem to be the right criterionin southeastern France, since it is related more to activitiesthan to settlement age. Since retouches are not often invasive,Dibble’s model cannot be applied in this case. Is the lowfrequency of flake tools in the upper level of Orgnac 3 linkedwith the extensive use of the Levallois method, providingflakes which did not need retouching or resharpening, andperhaps to the kinds of activities involved and the durationof use of the cutting edges?

The technical behaviours in the Middle Rhone Valley showgreat variability through time, and their meaning remains tobe interpreted precisely. Levallois flaking seems to be relatedto cold periods (Orgnac 3 OIS 8, Maras laminar method OIS4) and the discoid method with temperate ones (Payre OIS7-5, Saint-Marcel OIS 3). However, at sites such as Le Figuier(Quina) and Abri des Pecheurs a discoid method was usedin a cold context (OIS 4). Even though one method waspreferentially applied, others were also employed in all oc-cupations. At any given site, traditions and activities wereprobably mixed rather than time-related. Nevertheless, in thisarea, the cold climate opened up the landscape and made flintoutcrops more available. Severe conditions certainly influ-enced human mobility and subsistence activities and conse-quently technical choices.

In southeastern France, the transition toward Middle Pa-laeolithic behavior in the form of long and controlled flakeprocessing systems and standardized production appeared atthe OIS 9/8 border. The precocious disappearance of bifacesis assumed to indicate a slow process which began before OIS8 and ended around OIS 7-6-5. In all of western Europe, asat the southeastern French sites, artifacts such as bifaces andpebble tools are few; from the beginning, the tool kit is com-posed mainly of flake tools. When pebble tools are morenumerous, the assemblages are related neither to a specificperiod nor to a particular climatic context (Payre OIS 7-5,Abri des Pecheurs OIS 5-4), instead corresponding to specificactivities and needs.

In short, there are still many questions open for debate.Does site scarcity influence our interpretation of this period?Are our observations biased by the comparison with open-air sites on the great North European plains while southernsites were in caves in narrow basins and valleys? The transitionappears to have depended upon the geographic setting.

Lawrence Guy StrausDepartment of Anthropology, University of New Mexico,Albuquerque, NM 87131, U.S.A. ([email protected]). 30 III06

For a long time, most Paleolithic specialists have realized thatthe hoary nineteenth-century subdivisions of the Old StoneAge have unraveled even if many of the terms continue toserve as shorthand indicators of the approximate temporalplacement and broad technological characteristics of artifactassemblages. This is a very sensible and well-supported dis-cussion of diachronic change and synchronic variabilitywithin the “Early [Lower and Middle] Paleolithic.” Monnieris to be applauded for going back to unresolved issues ofsystematics and nomenclature in “paleo-lithic” studies—evenresurrecting application of the now-out-of-favor Bordesianmethod of assemblage description.

Monnier is correct to insist on the use of dating methodsindependent of the artifacts themselves; circularity must beavoided if at all possible, even if the “dates” are imprecise.This is the same argument that some (mostly Americans andBritons) made for the application of radiocarbon dating withregard to the Upper Paleolithic and even the so-called Middle/Upper Paleolithic transition decades ago, while the traditionalFrench view was to rely on lithic and osseous fossiles directeursand sedimentological correlations tied to the modified Alpineglacial chronostratigraphic framework (see Straus 1975, 1987,1991; Straus and Clark 1986). Freed from the slavish relianceon artifacts and artifact manufacturing techniques as essen-tially time-markers, we can approach an understanding ofinterassemblage variability in terms of raw-material avail-ability, site-specific place, and occupation function, as well assite formation/disturbance and archeological sampling effects.(Monnier hints at this in her perceptive comment on theunusually “crude” lithic assemblages associated with massesof bison bones at late Middle Paleolithic Le Mauran.

We are finally coming to grips with the consequences ofour profession’s longtime fealty to the restrictive and mis-leading de Mortilletian stages. Monnier does an excellent jobof explaining how these stage names persisted across changesin paradigms and in leading French prehistorians. Once itbecame clear, for example, that small bifacial handaxes couldbe found in the Mousterian, that many Acheulean-age assem-blages lack classic handaxes, and that the Levallois techniquewas invented in the Upper Acheulean, a simple bipartite di-vision of the European Early Paleolithic into Acheulean andMousterian became untenable, but much of the professionchose to ignore the fact. The increasing behavioral and tech-nical complexity of some Acheulean sites (e.g., the Schoningenwooden spears, the Bilzingsleben engraved bones) makes itclear that cultural variability and change were just as “mosaic”in the early–mid-Middle Pleistocene as they were during themid-Upper Pleistocene (see Straus 2005). True blade tech-


nology and bone points and even ritual or artistic expressionare now all acknowledged to be occasional aspects of theMiddle Paleolithic in Europe and in Africa, but we continueto think in terms of de Mortilletian stages as if the differencesbetween them constituted chasms between prehuman incom-petence and human genius. With little time in introductorycourses to devote to the vast span of early human evolutionor to nuanced explanations, we often find it necessary to resortto simplifications. Thus we fall into the trap of talking aboutthe characteristics of the Acheulean, the Mousterian, and theUpper Paleolithic even if we know how faulty such shortcutsactually are. We see how problematic the use of cultural“boxes” is when serious specialists talk of an Oldowan set-tlement of southern Europe long after the invention of thebifacial handaxe in Africa some 1.6 million years ago, evenas we know that chopper/chopping tool assemblages (a laBreuil’s parallel, nonbiface phylum) continue to pop up inthe record throughout much of the Middle Pleistocene andbeyond.

Although Monnier detects some new patterns that runcounter to the classic definitions of major periods in Homoheidelbergensis and H. sapiens neanderthalensis cultural de-velopment, we should be cautious about adopting them asnormative characterizations, since the sample of “well”-dated,recently excavated, and systematically analyzed sites used inthis study is rather small and may be unrepresentative of thefull range of human sites and activities in these vast spans oftime. As a “spoiler” argument, however, the case she makesis impressive.

Monnier’s excellent description of the de Mortillet-Com-mont-Breuil-Peyrony periodization schemes brings me backmemories of the artifact collection cabinet of my grandfather,Guy Magnant, in Bordeaux, the lower drawers of which arelabeled “Cheleen” and “Acheuleen” (Straus 1985). In the earlyyears of the discipline of prehistory, an instrument like thede Mortillet framework was necessary to bring some sem-blance of order to a confusing situation. Trading artifacts from“type sites” as they did, these early prehistorians had to“know” into which etage and literally which etagere to puttheir precious artifacts. But the idea of a rigid scheme of stageshas long since lost its utility with the development of chron-ometric dating methods and the increase in well-excavatedsites. It takes time to turn the Leviathan, but with careful,deliberate steering by helmsmen such as Monnier, the greatship does seem to be rounding a crucial buoy.

Alain TuffreauLaboratoire de Prehistoire et Quaternaire, Universite desSciences et Technologies de Lille, F-59655 Villeneuve d’Ascqcedex, France ([email protected]). 4 V 06

The periodization of the Paleolithic into three phases—Lower,Middle, and Upper—has regularly been called into question.Fundamentally, it depends upon the inclusion or absence of

specific characteristics of important lithic industries that havegradually appeared in the course of at least tens of thousandsof years. This does not, however, apply to the Upper Paleo-lithic, which developed rather rapidly after a period of tran-sition of some thousands of years. In the identification of theUpper Paleolithic, other characteristics are taken into con-sideration, especially stylistic features, a bone industry, andart.

Following a well-documented historical analysis of the per-iodization of the Lower/Middle Paleolithic, Monnier exam-ines whether the characteristics of tools justify a division be-tween the two. Her sample consists of 89 assemblages datedby radiometric methods to between OIS 17 and OIS 3. The29 layers from which the lithic assemblages originate are lo-cated in Western Europe. The purpose of the study is to provethat, although evolutionary tendencies are present, it is im-possible to determine a precise chronological boundary basedon the presence or absence of features or variations in theirfrequencies. Thus the percentage of bifaces tends to diminishafter OIS 8, whereas a significant presence of the Levalloistechnique is later (OIS 7) and retouched flakes are found ingreat quantities during the colder periods (OIS 6 and OIS 3).My observations will focus on the sample selected and thecriteria employed to differentiate the Lower from the MiddlePaleolithic.

The sample consists entirely of assemblages organized chro-nologically according to radiometric dates. The undertakingis fully justified for the karstic regions of southwestern Europe,where it is impossible to use chronostratigraphy to date lithicindustries. In the north of France the data have been con-siderably augmented in the past 30 years because of the studyof open-air deposits excavated during salvage operations. Thelack of preservation of bone remains and the rarity of burntflint are responsible for the scarcity of radiometric dates forthese deposits. Because they are found in a loess environment,however, they benefit from an excellent chronostratigraphicframework, one which, for the last glacial (OIS 5 to OIS 3),for example, allows them to be placed within a 10,000-yearrange (in other words, offers as much precision as radiometricdates). Including these lithic industries (Le Pucheuil andEtoutteville for OIS 8 to 6 and Bennecourt-Saint-Ouen, Rien-court-les-Bapaume, Seclin, and the deposits in Senonais forOIS 5 to OIS 3) would certainly have helped, as Monnierstresses, to reduce the overrepresentation of lithic industriesfrom southwestern France as far as the more recent periodsof time are concerned.

Central Europe is not included in the scope of this study,and, as a result, the Micoquian, a Middle Paleolithic industryrich in bifaces, has been all but ignored. These tools, like mostbifaces starting with OIS 8, are not comparable to Acheuleanbifaces sensu stricto, which tend to vanish in Western Europearound 280,000 BP. Many of them have retouched edges sim-ilar to those of flake tools. This means that they served asblanks and their function differed from that of Acheuleanbifaces. The bifacial tools of the Micoquian, similarly to those


from the Mousterian of Acheulean Tradition, also have ob-vious stylistic features. Finally, another characteristic that es-capes a study based entirely on deductions from index fossilsin terms of the Bordes method is the greater diversity ofmethods of knapping after OIS 7.

These comments do not in any way diminish the impor-tance of Monnier’s study. As she points out, the Lower/MiddlePaleolithic periodization is even more complex than her ar-ticle shows.1

Reply

I thank all of the commentators for the time and effort theyhave spent reading this article and commenting on it. Theymake important observations and critiques that I am happyto have a chance to address. I hope that the following responsewill clarify some misunderstandings related to the goals ofthis study and illuminate the complexity of many of the issuesI faced.

Tuffreau mentions that I should have included sites fromnorthern France such as Le Pucheuil, Etoutteville, Riencourt-les-Bapaume, and Seclin, which do not have absolute dates(because of a lack of suitable dating material) but haveexcellent chronostratigraphic control because of theirpreservation in loess sediments. McNabb and Fluck argue thatI should have included Barnham (which has problematicabsolute dates and is assigned to OIS 11 on the basis ofsedimentology and biostratigraphy [Ashton, Lewis, and Parfitt1998]), High Lodge (which does not have absolute dates),and Pontnewydd (which does not meet the criterion forminimum assemblage size), among others. As I explained, themain objective of my study was to test whether the diachronicpatterning of the tool types originally used to define the Lowerand Middle Paleolithic (bifaces, Levallois technology, andretouched flake tools) still supports the Lower/MiddlePaleolithic periodization. In order to achieve this objective,it was imperative that the dating of the assemblages I usedbe independent of artifact typology. Biostratigraphy andgeochronology are not entirely independent of typology (e.g.,some mammalian species are dated on the basis of thetypological composition of key sites). Absolute dates may be“imprecise,” in Straus’s words, but they should, in theory, beindependent of artifact typology. Therefore, while I wouldhave liked to augment my sample by including sites such asHigh Lodge, for which convincing geological arguments aremade that it is pre-Anglian (the Anglian is correlated withOIS 12), I had to be consistent in my application of selectioncriteria or risk losing objectivity.

As regards the geographic boundaries of the study, I agreewith Tuffreau that the inclusion of Micoquian and other

1. Translated by Simone Monnier Clay.

industries from Central Europe would be a valuable additionto this study. However, Central and Eastern Europe werebeyond the purview of this work, though they will certainlybe included in the future, along with the Near East and,eventually, Africa.

McNabb and Fluck point out that the dating of Hoxne andSwanscombe have been updated as a result of biostratigraphicaland sedimentological studies which place it in OIS 11 ratherthan OIS 9 (Schreve 2001 and Bridgland 1994). However, thiswould not substantially alter the results of my analysis. As canbe seen in table 8, I currently have no data from either OIS 10or 11. I would therefore update Hoxne and Swanscombe bychanging the category currently listed as “OIS 9” to a new oneconsisting of “OIS 9 through 11.” While a higher degree ofchronological resolution is desirable, at the present time thereare simply too few well-dated sites from this time period. Thenext few years may see a resolution of the dating problemssurrounding these sites.

Finally, regarding figure 9, McNabb and Fluck mistakenlystate that I include OIS 3 along with OIS 5.5, 7, 9, 11, 13.OIS 3 is included with OIS 4, 6, 8, and 12 and contrastedwith data from OIS 5, 7, and 9. In addition, they state thatOIS stages 5.1–5.4 are not interglacials and should not becompared with OIS 7 and 9; however, they do not representthe extreme glacial conditions of OIS 4–2, either.

Bisson and McNabb and Fluck object to my calculation ofartifact type frequencies relative to the total lithic material ofeach assemblage. They would prefer that I calculate them inthe traditional (Bordesian) way, which is relative to theretouched flake component only. As I have said, I used totallithic material as the denominator for several reasons. Firstof all, since the objective of the study was to trace thediachronic patterning of bifaces, retouched flake tools, andLevallois technology across the Lower and Middle Paleolithic,it was essential that any variation in flake tool versus coretool frequencies across this transition not confound the issuesbeing addressed. Second, the historic focus on certain artifacttypes and the neglect of others have reified the periodizationscheme and ignored other aspects of variability. For example,debitage, which is now known to contain valuable technologicaldata, was ignored for decades. What other aspects of thearchaeological record do we continue to ignore? Manuports?Broken cores? Shatter? Calculating artifact type frequenciesrelative to the total lithic assemblage shows us the truefrequencies of some supposedly key types (such as bifaces).Of course, taphonomic factors such as breakage will have hadan effect on the number of lithic pieces in a given assemblage.Thus, it would be more accurate to calculate artifactfrequencies on the basis of weights rather than counts.Unfortunately, data on weights are rarely available in theliterature. Third, there is little consistency among researcherswhen it comes to reporting artifact counts (e.g., inclusion orexclusion of broken implements), but they do all report totallithic materials. Fourth, McNabb and Fluck are concernedthat “site function, hominin transport behavior, excavation


history, resharpening and reuse” will affect debitage frequencies.They forget that the same factors also affect retouched toolfrequencies. Finally, inherent in McNabb and Fluck’s argumentsthat variable debitage frequencies can affect the “true statisticalimportance of significant tool types” (my italics) and Bisson’sstatement that there are “inherent differences in the ratios of‘finished products’ to unmodified flakes produced by differenttool-making strategies” is the assumption that twenty-first-century archaeologists are able to identify which tool types are“significant” and “finished.” Davidson and Noble (1993, 365)refer to this as the “finished artifact fallacy”—the “belief thatthe final form of flaked stone artefacts as found byarchaeologists was the intended shape of a ‘tool.”’ However,retouched flake tools are not necessarily finished, nor are theynecessarily more significant than unretouched blanks. Theirperceived significance is merely the product of an arbitrary,historical focus on certain retouched tools.

Bisson argues that biface frequencies should not becalculated relative to total lithic material, stating that themanufacture of a biface produces more debitage than themanufacture of a unifacial flake tool such as a scraper. As anexample, he calculates that French Acheulean bifaces have anaverage of 20 flake scars each [over 2 cm], while FrenchMiddle Paleolithic scrapers have an average of only 3.9 scarseach. Bifaces, however, are core tools; therefore, they willsurely have many more flake removals than flake tools, onlythe edges of which are retouched. Furthermore, bifaces havetwo faces, which approximately doubles the number of flakescars relative to unifacial retouched flake tools. A moreappropriate comparison would be the number of flakes createdduring the manufacture of a biface versus the number of flakescreated during the reduction of a core from which a flake wasthen selected and retouched.

McNabb and Fluck are concerned by my use of Bordesiantypology, which they say does not apply to Britain, and Strausremarks that I “even resurrect application of the now-out-of-favor Bordesian method of assemblage description.” Tuffreausuggests that I should have taken into account the greaterdiversity of knapping methods after OIS 7. Moncel providesa useful summary of technological trends in southeasternFrance from OIS 8 to OIS 3, concluding that while someclimatic trends may be present, traditions and activitiestranscend chronology. I agree with these commentators thatthe Bordes method of assemblage analysis is old-fashioned(although not useless, by any means) and that studies of lithictechnology, specifically core reduction sequences, are valuablesources of information for answering many questions inPaleolithic studies, not least of which are those regardingchanges through time. However, the objective of my study wasnot to explore all aspects of change in lithic assemblagesthroughout the Lower and Middle Paleolithic; it was merely totest the diachronic patterning of the index fossils defined overa century ago and still used to characterize the Lower andMiddle Paleolithic. Since my focus was on traditional tool types(bifaces, flake tools, Levallois flakes), I decided to use Bordes’s

typology, which was designed for western Europe and has beenwidely used there for decades. I agree with McNabb and Fluckthat there may be tool types unique to the British LowerPaleolithic that deserve special treatment, but the categories Iused, such as “retouched flake tools” and “bifaces,” werepurposefully designed to be broad and to address the specificquestions of the study, not to document the idiosyncratic natureof each regional sequence. Regarding core reduction methods,I believe that syntheses of the diachronic patterning of debitagemethods such as Levallois, discoidal, laminar, etc., will be veryvaluable once questions regarding methodology (see, e.g.,Tostevin 1996) have been addressed.

—Gilliane Monnier

References Cited

Aldhouse-Green, S. 1998. The archaeology of distance: Per-spectives from the Welsh Palaeolithic. In Stone Age ar-chaeology: Essays in honour of John Wymer, ed. N. Ashton,F. Healy, and P. Pettitt, 137–415. Lithic Studies Society Oc-casional Paper 6. [JM, HF]

———. 2001. The colonisations and visitations of Wales byNeanderhals and modern humans in the Later Pleistocene.In Histories of old ages: Essays in honour of Rhys Jones, ed.A. Anderson, I. Lilley, and S. O’Connor, 225–35. Canberra:Pandus Books. [JM, HF]

Anconetani, P., C. Crovetto, M. Ferrari, G. Giusberti, L.Longo, C. Peretto, and F. Vianello. 1992. Nuove ricerchenel giacimento di Isernia La Pineta (Molise). Rivista diScienze Preistoriche 44.

Antoine, Pierre. 2001. La stratigraphie de la sequence fluviatilede Cagny-la-Garenne II (Somme). In L’Acheuleen dans lavallee de la Somme: Donnees recentes, ed. A. Tuffreau, 45–50.Lille: Universite des Sciences et Technologies de Lille.

Ashton, N., J. Cook, and J. Rose, eds. 1992. High Lodge: Ex-cavations by G. de G. Sieveking 1962–1968 and J. Cook 1988.London: British Museum Press. [JM, HF]

Ashton, N., P. Dean, and J. McNabb. 1991. Flaked flakes:What, where, when and why? Lithics 12:1–11. [JM, HF]

Ashton, Nick, Brian Irving, Simon Lewis, Simon Parfitt, andJohn McNabb. 1994. Contemporaneity of Clactonian andAcheulian flint industries at Barnham, Suffolk. Antiquity68:585–89.

Ashton, N., S. G. Lewis, and S. Parfitt, eds. 1998. Excavationsat the Lower Palaeolithic site at East Farm, Barnham,1989–1994. British Museum Occasional Paper 125 [JM, HF]

Ashton, N., S. G. Lewis, S. Parfitt, I. Candy, D. H. Keen, R.Kemp, K. Penckman, G. Thomas, J. Whittaker, and M. J.White. 2005. Excavations at the Lower Palaeolithic site atElveden, Suffolk, UK. Proceedings of the Prehistoric Society71:1–61. [JM, HF]

Ashton, Nick, and John McNabb. 1992. The interpretationand context of the High Lodge industries. In High Lodge:


Excavations by G. de G. Sieveking 1962–8 and J. Cook 1988,ed. N. Ashton, J. Cook, S. G. Lewis, and J. Rose, 164–68.London: British Museum Press.

———. 1996. The flint industries from the Waechter exca-vations. In Excavations at Barnfield Pit, Swanscombe,1968–72, ed. B. Conway, J. McNabb, and N. Ashton,201–36. British Museum Occasional Paper 94.

Bahain, Jean-Jacques, Christophe Falgueres, and Michel Lau-rent. 2001. Datation par resonance paramagnetique elec-tronique (RPE) de sediments et par combination des meth-odes RPE/U-Th de restes paleontologiques (provenant dessites paleolithiques de Cagny). In L’Acheuleen dans la valleede la Somme: Donnees recentes, ed. A. Tuffreau, 37–40. Lille:Universite des Sciences et Technologies de Lille.

Bar-Yosef, Ofer. 1982. Some remarks on the nature of tran-sitions in prehistory. In The transition from Lower to MiddlePalaeolithic and the origin of modern man, ed. A. Ronen,29–33. British Archaeological Reports International Series151.

———. 1995. The Lower and Middle Paleolithic in the Med-iterranean Levant: Chronology and cultural entities. In Manand environment in the Palaeolithic, ed. H. Ullrich, 246–63.Liege: Universite de Liege. [RGK]

Bisson, Michael S. 2000. Nineteenth-century tools for twenty-first century archaeology? Why the Middle Paleolithic ty-pology of Francois Bordes must be replaced. Journal ofArchaeological Method and Theory 7(1):1–48. [MSB]

Binford, Lewis. 1973. Interassemblage variability: The Mous-terian and the functional argument. In The explanation ofculture change, ed. C. Renfrew, 227–54. London:Duckworth.

Binford, Lewis, and Sally Binford. 1966. A preliminary analysisof functional variability in the Mousterian of Levallois fa-cies. American Anthropologist 68:238–95.

Blackwell, Bonnie, and Henry Schwarcz. 1988. Datations desspeleothems de la grotte Vaufrey par la famille de l’uranium.In La Grotte Vaufrey: Paleoenvironnement, chronologie, ac-tivites humaines, ed. Jean-Philippe Rigaud, 365–80. Me-moires de la Societe Prehistorique Francaise 19.

Blackwell, Bonnie, Henry Schwarcz, and Andre Debenath.1983. Absolute dating of hominids and Paleolithic artefactsof the cave of La Chaise-de-Vouthon (Charente), France.Journal of Archeological Science 10:493–513.

Boeda, Eric, Jean-Michel Geneste, and Lilianne Meignen.1990. Identification de chaınes operatoires lithiques du Pa-leolithique ancien et moyen. Paleo 2:43–80.

Bonifay, E., and B. Vandermeersch. 1991. Les premiers Eu-ropeens. Paris: Editions du Comite des Travaux Historiqueset Scientifiques.

Bordes, Francois. 1950a. Principes d’une methode d’etude destechniques de debitage et de la typologie du Paleolithiqueancien et moyen. L’Anthropologie 54:19–34.

———. 1950b. L’evolution buissonnante des industries enEurope Occidentale: Considerations theoriques sur le Pa-leolithique ancien et moyen. L’Anthropologie 54:393–420.

———. 1954. Notules de typologie paleolithique. 3. Pointesmousteriennes, racloirs convergents et dejetes, limaces. Bul-letin de la Societe Prehistorique Francaise 51:336–39.

———. 1961a. Typologie du Paleolithique ancien et moyen.Paris: Centre National de la Recherche Scientifique.

———. 1961b Mousterian cultures in France. Science 134:803–10.

———. 1969. Reflections on typology and techniques in thePaleolithic. Arctic Anthropology 6:1–29.

———. 1970. Reflexions sur l’outil au Paleolithique. Bulletinde la Societe Prehistorique Francaise 67:199–202.

———. 1972. A tale of two caves. New York: Harper and Row.Bordes, Francois, and Maurice Bourgon. 1951. Le complexe

mousterien: Mousteriens, levalloisien et tayacien.L’Anthropologie 55:1–23.

Bordes, Francois, Jean-Philippe Rigaud, and Denise de Son-neville-Bordes. 1972. Des buts, problemes et limites del’archeologie paleolithique. Quaternaria 16:15–34.

Bordes, Francois, and Denise de Sonneville-Bordes. 1970. Thesignificance of variability in Paleolithic assemblages. WorldArchaeology 2:61–73.

Bosinski, Gerhardt. 1982. The transition Lower/Middle Pa-laeolithic in northwestern Germany. In The transition fromLower to Middle Palaeolithic and the origin of modern man,ed. A. Ronen, 165–175. British Archaeological Reports In-ternational Series 151.

———. 1995. Stone artefacts of the European Lower Pa-laeolithic: A short note. In The earliest occupation of Europe,ed. Wil Roebroeks and Thijs van Kolfschoten, 263–68. Lei-den: University of Leiden.

———. 1996. Atlas des sites du Paleolithique inferieur. Paris:Errance.

Bowen, D. Q., S. Hughes, G. A. Sykes, and G. H. Miller. 1989.Land-sea correlations in the Pleistocene based on isoleucineepimerization in non-marine molluscs. Nature 340:49–51.[JM, HF]

Bowman, Sheridan. 1993. Themoluminescence dating of theLower Industry. In The Lower Paleolithic site at Hoxne,England, ed. R. Singer, B. G. Gladfelter, and J. J. Wymer,208–10. Chicago: University of Chicago Press.

Breuil, Henri. 1932a Le paleolithique ancien en Europe oc-cidentale et sa chronologie. Bulletin de la Societe Prehisto-rique Francaise 29:570–78.

———. 1932b. Les industries a eclat du Paleolithique ancien.1. Le Clactonien. Prehistoire 1:125–90.

Breuil, Henri, and L. Koslowski. 1931. Etudes de stratigraphiepaleolithique dans le nord de la France, la Belgique etl’Angleterre. L’Anthropologie 41:449–88.

———. 1932a. Etudes de stratigraphie paleolithique dans lenord de la France, la Belgique, et l’Angleterre. L’Anthropologie42:27–47.

———. 1932b. Etudes de stratigraphie paleolithique dans lenord de la France, la Belgique et l’Angleterre. L’Anthropologie42:291–314.

———. 1934a. Etudes de stratigraphie paleolithique dans le


nord de la France, la Belgique et l’Angleterre. L’Anthropologie44:249–90.

———. 1934b. Etudes de stratigraphie paleolithique dans lenord de la France, la Belgique, et l’Angleterre. L’Anthropologie44:449–88.

Bridgland, D. R. 1994. Quaternary of the Thames. London:Chapman and Hall. [JM, HF]

Bridgland, D. R., M. H. Field, J. A. Holmes, J. McNabb, R.C. Preece, I. Selby, J. Wymer, S. Boreham, B. Irving, S. A.Parfitt, and A. J. Stuart. 1999. Middle Pleistocene intergla-cial Thames-Medway deposits at Clacton-on-Sea, England:Reconsideration of the biostratigraphical and environmen-tal context of the type Clactonian Palaeolithic Industry.Quaternary Science Reviews 18:109–46. [JM, HF]

Bridgland, D., P. Gibbard, P. Harding, R. A. Kemp, and G.Southgate. 1985. New information and results from recentexcavations at Barnfield Pit, Swanscombe. QuaternaryNewsletter 46:25–39.

Callow, Paul, and J. Cornford, eds. 1986. La Cotte de St.Brelade 1961–1978: Excavations by C. B. M. McBurney. Nor-wich: Geo Books.

Capitan, L., and Denis Peyrony. 1912. Station prehistoriquede la Ferrassie, commune de Savignac-du-Bugue (Dor-dogne). Revue Anthropologique 22:29–50, 76–99.

Carboneıl, Eudald, M. Garcla-Anton, Carolina Mallol, M.Mosquera, A. Olle, X. P. Rodrıguez, Mohamed Sahnouni,Robert Sala, and J. M. Verges. 1999. The TD6 level lithicindustry from Gran Dolina, Atapuerca (Burgos, Spain):Production and use. Journal of Human Evolution 37:653–93.

Carbonell, Eudald, and X. Rodriguez. 1994. Early MiddlePleistocene deposits and artefacts in the Gran Dolina Site(TD4) of the “Sierra de Atapuerca” (Burgos, Spain). Journalof Human Evolution 26:291–312.

Clark, Grahame. 1969. World prehistory: A new outline. 2d ed.Cambridge: Cambridge University Press.

Clark, J. D., and K. B. Schick. 2000. Overview and conclusionon the Middle Awash Acheulean. In The Acheulean and thePlio-Pleistocene deposits of the Middle Awash Valley, Ethiopia,ed. J. de Heinzelin, J. D. Clark, K. B. Schick, and H. Gilbert,193–202. Tervuren: Musee Royal de l’Afrique Centrale.[RGK]

Collins, Desmond. 1970. Stone artefact analysis and the rec-ognition of culture traditions. World Archaeology 2:17–27.

Commont, Victor. 1906. Les decouvertes recentes a Saint-Acheul: L’Acheuleen. Revue de l’Ecole d’Anthropologie deParis 16:228–41.

———. 1908. Les industries de l’ancien Saint-Acheul.L’Anthropologie 19:527–72.

———. 1913. Les hommes contemporains du Renne dans lavallee de la Somme. Amiens: Imprimerie Yvert and Tellier.

Conard, N. J., and B. Fischer. 2000. Are there recognizablecultural entities in the German Middle Palaeolithic? In To-ward modern humans: Yabrudian and Micoquian, 400–50 kyears ago, ed. A. Ronen and M. Weinstein-Evron, 7–24.

British Archaeological Reports International Series S850.[RGK]

Conway, Bernard. 1995. Bifaces in a Clactonian context atLittle Thurrock, Grays, Essex. Lithics 16:41–46.

———. 1996. The stratigraphy and chronology of the Pleis-tocene deposits of Barnfield Pit, Swanscombe. In Excava-tions at Barnfield Pit, Swanscombe, 1968–72, ed. B. Conway,J. McNabb, and N. Ashton, 117–36. British Museum Oc-casional Papers 94.

Conway, Bernard, John McNabb, and Nick Ashton, eds. 1996.Excavations at Barnfield Pit, Swanscombe, 1968–72. BritishMuseum Occasional Papers 94.

Cremaschi, M., and C. Peretto. 1988. Les sols d’habitat dusite paleolithique d’Isernia la Pineta (Molise, Italie cen-trale). L’Anthropologie 92:1017–40.

Darlas, H. 1994. L’Acheuleen final des couches superieuresde la grotte du Lazaret (Nice-Alpes-Maritimes). L’Anthro-pologie 92:267–304.

Davidson, Ian, and William Noble. 1993. Tools and languagein human evolution. In Tools, language, and cognition inhuman evolution, ed. K. Gibson and T. Ingold, 363–88.Cambridge: Cambridge University Press.

Debard, E. 1988. Le Quaternaire du Bas-Vivarais: Dynamiquesedimentaire, paleoclimatique et chronologie des remplissagesd’avens, de porches, de grottes et d’abris sous roche; compa-raisons avec le Velay oriental. Lyon: Laboratoire de Geologiede Lyon.

Debard, E., and J. F. Pastre. 1988. Un marqueur chronostra-tigraphique du Pleistocene moyen a la peripherie du MassifCentral: La retombee a clinopyroxene vert du Sancy dansle site acheuleen d’Orgnac III (Bas-Vivarais-S.-E. France).Comptes Rendus de l’Academie des Sciences 306:1515–20.

Debenath, Andre. 1974. Recherches sur les terrains Quater-naires charentais et les industries qui leur sont associees.Ph.D. diss., Universite de Bordeaux.

Delagnes, Anne. 1992. L’organisation de la production li-thique au Paleolithique Moyen: Approche technologique apartir de l’etude des industries de la Chaise-de-Vouthon(Charente). Ph.D., diss., Universite de Paris X.

Delpech, Francoise, and Henri Laville. 1988. Climatologie etchronologie de la grotte Vaufrey: Confrontation des hy-potheses et implications. In La Grotte Vaufrey: Paleoenvi-ronnement, chronologie, activites humaines, ed. J.-Ph. Ri-gaud, 381-88. Memoires de la Societe Prehistorique Fran-caise 19.

Dennell, Robin. 2003. Dispersal and colonisation, long andshort chronologies: How continuous is the Early Pleisto-cene record for hominids outside East Africa? Journal ofHuman Evolution 45:421–40.

Dennell, Robin, and Wil Roebroeks. 1996. The earliest col-onization of Europe: The short chronology revisited. An-tiquity 70:535–42.

d’Errico, Francesco. 2003. The invisible frontier:A multiplespecies model for the origin of behavioral modernity. Evo-lutionary Anthropology 12:188–202.


d’Errico, Francesco, J. Zilhao, D. Baffier, M. Julien, and J.Pelegrin. 1998. Neandertal acculturation in western Eu-rope? A critical review of the evidence and its interpretation.Current Anthropology 39:S1–S44.

Despriee, J., R. Gageonnet, P. Voinchet, J.-J. Bahain, C. Fal-gueres, and J. Depont. 2005. Les industries a bifaces dansles formations fluviatiles du bassin de la Loire (Region Cen-tre, France). In Les premiers peuplements en Europe (Rennes,22–25 septembre 2003), ed. N. Molines, M-H. Moncel, andJ-L. Monnier, 431–44. British Archaeological Reports In-ternational Series S1364. [MHM]

Dibble, Harold L. 1987. The interpretation of Middle Pale-olithic scraper morphology. American Antiquity 52:109–17.

———. 1991. Local raw material exploitation and its effectson Lower and Middle Paleolithic assemblage variability. InRaw material economies among prehistoric hunter-gatherers,ed. A. Montet-White and S. Holen, 33–48. University ofKansas Publications in Anthropology 19.

———. 1995. Middle Paleolithic scraper reduction: Back-ground, clarification, and review of evidence to data. Jour-nal of Archaeological Method and Theory 2:299–368.

Dibble, Harold L., and Michel Lenoir, eds. 1995. The MiddlePaleolithic site of Combe-Capelle Bas (France). Philadelphia:University Museum Press.

Dibble, Harold L., and Nicolas Rolland. 1992. On assemblagevariability in the Middle Paleolithic of Western Europe:History, perspectives, and a new synthesis. In The MiddlePaleolithic: Adaptation, behavior, and variability, ed. H. Dib-ble and P. Me1lars, 1–20. Philadelphia: University MuseumPress.

Falgueres, C., J. J. Bahain, Y. Yokoyama, J. L. Arsuaga, J. M.Bermudez de Castro, Eudald Carbonell, J. L. Bischoff, andJ. M. Dolo. 1999. Earliest humans in Europe: The age ofTD6 Gran Dolina, Atapuerca, Spain. Journal of HumanEvolution 37:343–52.

Falgueres, C., G. Shen, and Y. Yokoyama. 1988. Datation del’aven d’Orgnac III: Comparaison par les methodes de laresonnance de spin electronique (ESR) et du desequilibredes familles de l’uranium. L’Anthropologie 92:727–30.

Freeman, Leslie G. 1996. The nature of Mousterian facies inCantabrian Spain. In Recent studies in paleoanthropology,ed. J. D Clark and F. C. Howell, 230–37. American An-thropologist n.s., 68, no. 2, p. 2.

Gagnepain, J., and C. Gaillard. 2005. La grotte de la BaumeBonne (Quinson, Alpes de Haute-Provence): Synthesechronostratigraphique et sequence culturelle d’apres lesfouilles recentes (1988–1997). In Les premiers peuplementsen Europe (Rennes, 22–25 septembre 2003), ed. N. Molines,M-H. Moncel, and J-L. Monnier, 73–87. British Archaeo-logical Reports International Series S1364. [MHM]

Gamble, Clive. 1999. The Palaeolithic societies of Europe. Cam-bridge: Cambridge University Press.

Gamble, Clive, and Gilbert Marshall. 2001. The shape ofhandaxes, the structure of the Acheulian world. In A veryremote period indeed: Papers on the Palaeolithic presented to

Derek Roe, ed. Sarah Milliken and Jill Cook. Oxford: OxbowBooks.

Gamble, Clive, and Wil Roebroeks. 1999. The MiddlePaleolithic: A point of inflection. In The Middle Palaeolithicoccupation of Europe, ed. W. Roebroeks and C. Gamble,3–22. Leiden: University of Leiden.

Gao, Xing, and Christopher J. Norton. 2002. A critique ofthe Chinese “Middle Paleolithic.” Antiquity 76:397–412.

Geneste, Jean-Michel. 1985. Analyse d’industries mousterien-nes du Perigord: Une approche technologique da com-portement des groupes humains au Paleolithique moyen.Ph.D. diss., Universite de Bordeaux.

Groenen, Marc. 1994. Pour une histoire de la prehistoire. Gre-noble: Jerome Millon.

Grun, Rainer, Paul Mellars, and Henri Laville. 1991. ESR chro-nology of a 100,000-year archaeological sequence at Pechde l’Aze II, France. Antiquity 65:544–51.

Hayden, B. 1993. The cultural capacities of Neandertals: Areview and re-evaluation. Journal of Human Evolution 24:113–46.

Henshilwood, C. S., F. d’Errico, C. W. Marean, R G. Milo,and R. Yates. 2001. An early bone tool industry from theMiddle Stone Age at Blombos Cave, South Africa: Impli-cations for the origins of modern human behaviour, sym-bolism, and language. Journal of Human Evolution 41:631-78.

Hovers, Erella, and Steven L. Kuhn, eds. 2005. Transitionsbefore the transition: Evolution and stability in the MiddlePaleolithic and Middle Stone Age. New York: Kluwer Press.

Huxtable, J. 1986. The thermoluminescence dates. In La Cottede St. Brelade 1961–1978: Excavations by C. B. M. McBurney,ed. P. Callow and J. Cornford, 145–49. Norwich: GeoBooks.

Isaac, G. L. 1975. Stratigraphy and cultural patterns in EastAfrica during the middle ranges of Pleistocene time. InAfter the australopithecines, ed. K. W. Butzer and G. L. Isaac,543–69. The Hague: Mouton. [RGK]

Jaubert, Jacques. 1993. Le gisement paleolithique moyen deMauran (Haute-Garonne): Techno-economie des indus-tries lithiques. Bulletin de la Societe Prehistorique Francaise90:328–35.

Jaubert, Jacques, M. Lorblanchet, Henri Laville, R. Slott-Moller, Alain Turq, and Jean-Philippe Brugal. 1990. Leschasseurs d’aurochs de La Borde: Un site du PaleolithiqueMoyen (Livernon, Lot). Paris: Editions de 1a Maison desSciences de l’Homme.

Jelinek, Arthur J. 1977. The Lower Paleolithic: Current evi-dence and interpretations. Annual Review of Anthropology6:11–32.

———. 1982. The Middle Paleolithic in the Southern Levant,with comments on the appearance of modern Homo sap-iens. In The transition from the Lower to Middle Paleolithicand the origin of modern man, ed. A. Ronen, 57–101. BritishArchaeological Reports International Series 151.

Klein, Richard G. 1999. The human career: Human biological


and cultural origins. 2d ed. Chicago: University of ChicagoPress.

Klein, R. G., and B. Edgar. 2002. The dawn of human culture.New York: Wiley. [RGK]

Kuhn, Steven L. 1995. Mousterian lithic technology. Princeton:Princeton University Press.

Kuhn, Steven L., and Erella Hovers. 2005. General introduc-tion. In Transitions before the transition: Evolution and sta-bility in the Middle Paleolithic and Middle Stone Age, ed. E.Hovers and S. Kuhn. New York: Kluwer Press.

Lamotte, Agnes, and Alain Tuffreau. 2001a. Le gisement deCagny-la-Garenne (Somme, France): Les industries li-thiques de Cagny-la-Garenne II (Somme, France). InL’Acheuleen dans la vallee de la Somme: Donnees recentes, ed.A. Tuffreau, 59–90. Lille: Universite des Sciences et Tech-nologies de Lille.

———. 2001b. Les industries lithiques de la sequence flu-viatile fine de Cagny-l’Epinette (Somme). In L’Acheuleendans la vallee de la Somme: Donnees recentes, ed. A. Tuffreau,113–34. Lille: Universite des Sciences et Technologies deLille.

Lartet, E., and H. Christy. 1865–75. Reliquiae Aquitanicae,being contributions to the archaeology and palaeontology ofPerigord and the adjoining provinces of southern France. Lon-don: Williams and Norgate.

Laville, Henri. 1973. The relative position of Mousterian in-dustries in the climatic chronology of the early Wurm inthe Perigord. World Archaeology 4:321–29.

———. 1982. On the transition from “Lower” to “Middle”Palaeolithic in south-west France. In The transition fromLower to Middle Palaeolithic and the origin of modern man,ed. A. Ronen, 131–35. British Archaeological Reports In-ternational Series 151.

Leroi-Gourhan, A. 1943. Evolution et techniques: L’homme etla matiere. Paris: Albin Michel.

———. 1964. Le geste et la parole. Vol. 1. Techniques et langage.Paris: Albin Michel.

Lubbock, J. 1865. Pre-historic times, as illustrated by ancientremains, and the manners and customs of modern savages.London: Williams and Norgate.

McBrearty, Sally, and Alison S. Brooks. 2000. The revolutionthat wasn’t: A new interpretation of the origin of modernhuman behavior. Journal of Human Evolution 39:453–563.

McNabb, John. 1996. More from the cutting edge: Furtherdiscoveries of Clactonian bifaces. Antiquity 70:428–36.

McPherron, A., and V. Schmidt. 1983. Paleomagnetic datingat Isernia la Pineta. In Isernia La Pineta, un accampamentopiu antico di 700,000 anni, ed. C. Peretto, C. Terzani, andM Cremaschi, 67–69. Bologna: Calderini.

McPherron, Shannon. 2000. Handaxes as a measure of themental capabilities of early hominids. Journal of Archaeo-logical Science 27:655–63.

Meignen, Liliane. 1993. L’Abri des Canalettes: Un habitat mous-terien sur les Grands Causses (Nant, Aveyron), fouilles

1980–1986. Paris: Centre National de la Recherche Scien-tifique.

Mellars, Paul. 1965. Sequence and devlopment of the Mous-terian traditions in southwestern France. Nature 205:626–27.

———. 1969. The chronology of Mousterian industries inthe Perigord region. Proceedings of the Prehistoric Society35:134–71.

———. 1996. The Neanderthal legacy. An archaeological per-spective from Western Europe. Princeton: Princeton Uni-versity Press.

Milliken, Sarah. 2001. Acheulian handaxe variability in Mid-dle Pleistocene Italy: A case study. In A very remote periodindeed: Papers on the Palaeolithic presented to Derek Roe,ed. Sarah Milliken and Jill Cook, 160–73. Oxford: OxbowBooks.

Moncel, Marie-Helene. 1999. Les assemblages lithiques du sitepleistocene moyen d’Orgnac 3 (Ardeche, Moyenne Vallee duRhone, France). Liege: Etudes et Recherches Archeologiquesde l’Universite de Liege.

———. 2003. L’exploitation de l’espace et la mobilite des grou-pes humains au travers des assemblages lithiques a la fin duPleistocene moyen et au debut du Pleistocene superieur: Lamoyenne vallee du Rhone entre Drome et Ardeche. BritishArcheological Repots International Series S1184. [MHM]

Moncel, Marie-Helene, and Jean Combier. 1992a. Industrielithique du site pleıstocene moyen d’Orgnac 3 (Ardeche).Gallia Prehistoire 34:1–55.

———. 1992b. Outillage sur eclat dans l’industrie lithique dusite pleistocene moyen d’Orgnac 3 (Ardeche, France).L’Anthropologie 96:5–48.

Monnier, Gilliane. 2000. A re-evaluation of the archaeologicalevidence for a Lower/Middle Paleolithic division in WesternEurope. Ph.D. diss., University of Pennsylvania.

———. 2006. Testing retouched flake tool standardizationduring the Middle Paleolithic: Patterns and implications.In Transitions before the transition: Evolution and stabilityin the Middle Paleolithic and Middle Stone Age, ed. E. Hoversand S. L. Kuhn. New York: Kluwer Press.

Mortillet, Gabriel de. 1867. Promenades prehistoriques al’exposition universelle. Materiaux pour Servir a l’HistoirePositive et Philosophique de l’Homme 3:181–283, 285–368.

———. 1869. Essai d’une classification des cavernes et desstations sous abri, fondee sur les produits de l’industriehumaine. Materiaux pour Servir a l’Histoire Primitive del’Homme 5:172–79.

———. 1873. Classification des diverses periodes de l’Age dela Pierre. In Congres International d’Anthropologie etd’Archeologie Prehistoriques, 6eme session, 432–59.

——— 1883. Le prehistorique: Antiquite de l’homme. Paris:C. Reinwald.

Mortillet, Gabriel de, and Adrien de Mortillet. 1900. Le pre-historique: Origine et antiquite de l’homme. 3d ed. Paris: C.Reinwald, Schleicher Freres.


———. 1903. Musee prehistorique. 2d ed. Paris: Reinwald,Schleicher Freres.

Movius, Hallam. 1948. The Lower Pa1eolithic cultures ofsouthern and eastern Asia. Transactions of the AmericanPhilosophical Society 38:329–420.

———. 1949. Lower Paleolithic archaeology in southern andeastern Asia. Studies in Physical Anthropology 1:17–81.

Mussi, Margherita. 1995. The earliest occupation of Europe:Italy. In The earliest occupation of Europe, ed. W. Roebroeksand T. van Kolfschoten, 27–49. Leiden: University ofLeiden.

Ohel, Milia Y. 1979. The Clactonian: An independent complexor an integral part of the Acheulean? Current Anthropology20:685–726.

Pares, J. M., and A. Perez-Gonzalez. 1999. Magnetochro-nology and stratigraphy at Gran Dolina section, Atapuerca(Burgos, Spain). Journal of Human Evolution 37:325–42.

Parfitt, S. et al. 2005. The earliest record of human activityin northern Europe. Nature 438:1008–12.

Pelegrin, J. 1995. Technologie lithique: Le Chatelperronien deRoc-de-Combe (Lot) et de la Cote (Dordogne). Paris: CNRS.

Peretto, C., ed. 1991 Isernia la Pineta, nuovi contributi scien-tifici. Isernia: Instituto Regionale per gli Storici del Molise“V. Cuoco.”

Peretto, C., C. Terzani, and M Cremaschi, eds. 1983. IserniaLa Pineta, un accampamento piu antico di 700.000 anni.Bologna: Calderini.

Peyrony, Denis. 1920. Le Mousterien—ses facies. AssociationFrancaise pour l’Avancement des Sciences 44:496–97.

———. 1930. Le Moustier: Ses gisements, ses industries, sescouches geologiques. Revue Anthropologique 40:1–50.

———. 1934a. Stratigraphie du gisement prehistorique deCombe-Capelle, Dordogne. In Congres Prehistorique deFrance, 2eme session, Perigueux, 418–20.

———. 1934b. La Ferrassie: Mousterien, Perigordien, Aurig-nacian. Prehistoire 3:1–92.

———. 1938. La Micoque, les fouilles recentes, leurs signi-fications. Bulletin de la Societe Prehistorique de Francaise 6:257–88.

———. 1943. Combe-Capelle. Bulletin de la Societe Prehis-tique Francaise 40:243–55.

Quade, J., N. Levin, S. Semaw, D. Stout, P. R. Renne, M. J.Rogers, and S. W. Simpson. 2004. Paleoenvironments ofthe earliest stone toolmakers, Gona, Ethiopia. Bulletin ofthe Geological Society of America 116:1529–44. [RGK]

Rigaud, Jean-Philippe, ed. 1988. La Grotte Vaufrey: Paleoen-vironnement chronologie, activites humaines. Memoires dela Societe Prehistorique Francaise 19.

Roberts, Mark B. 1986. Excavation of the Lower Paleolithicsite at Amey’s Eartham Pit, Boxgrove, West Sussex: A pre-liminary report. Proceedings of the Prehistoric Society 52:215–45.

Roberts, Mark B., Clive S. Gamble, and David R. Bridgland.1995. The earliest occupation of Europe: The British Isles.In The earliest occupalion of Europe, ed. Wil Roebroeks and

Thijs van Kolfschoten, 165–92. Leiden: University ofLeiden.

Roberts, Mark B., and Simon A. Parfitt, eds. 1999a. Boxgrove:A Middle Pleistocene hominid site at Eartham Quarry,Boxgrove, West Sussex. English Heritage Archaeological Re-port 17.

———. 1999b. Biostratigraphy and summary. In Boxgrove:A Middle Pleistocene hominid site at Eartham Quarry,Boxgrove, West Sussex, ed. M. B. Roberts and S. A. Parfitt,303–8. English Heritage Archaeological Report 17.

Roberts, Mark, Christopher Stringer, and S. Parfitt. 1994. Ahominid tibia from Middle Pleistocene sediments atBoxgrove, UK. Nature 369:311–13.

Roche, H., J.-P. Brugal, A. Delagnes, C. S. Feibel, S. Harmand,M. Kibunjia, S. Prat, and P.-J. Texier. 2003. Les sites ar-cheologiques plio-pleistocenes de la formation de Nachu-kui, Ouest-Turkana, Kenya: Bilan synthetique 1997–2001.Comptes Rendus Paleovol 2:663–73. [RGK]

Roe, D. A. 1982. Transition from Lower to Middle Palaeo-lithic, with particular reference to Britain. In The transitionfrom Lower to Middle Paleolithic and the origin of modernhumans, ed. A. Ronen, 177–90. British Archaeological Re-ports International Series 151.

Roebroeks, Wil. 1988. From find scatters to early hominid be-haviour: A study of Middle Palaeolithic riverside settlementsat Maastricht-Belvedere, The Netherlands. Leiden: Univer-sity of Leiden.

Roebroeks, Wil, and Raymond Corbey. 2001. Biases and dou-ble standards in palaeoanthropology. In Studying humanorigins: Disciplinary history and epistemology, ed. RaymondCorbey and Wil Roebroeks, 67–76. Amsterdam: Amster-dam University Press.

Roebroeks, Wil, and Clive Gamble, eds. 1999. The MiddlePalaeolithic occupation of Europe. Leiden: Universlty ofLeiden.

Roebroeks, Wil, and M. Mussi. 2000. Hunters of the GoldenAge: The Mid Upper Paleolithic of Eurasia 30,000–20,000Bp. Leiden: University of Leiden.

Roebroeks, Wil, and Alain Tuffreau. 1999. Palaeoenvironmentand settlement patterns of the Northwest European MiddlePaleolithic. In The Middle Palaeolithic occupation of Europe,ed. W. Roebroeks and C. Gamble, 3–22. Leiden: Universityof Leiden.

Roebroeks, Wil, and Thijs van Kolfschoten, ed. 1995a. Theearliest occupation of Europe: Proceedings of the EuropeanScience Foundation Workshop at Tautavel (France), 1993.Leiden: Univerity of Leiden.

———. 1995b. The earliest occupation of Europe: A reap-praisal of artefactual and chronological evidence. In Theearliest occupation of Europe, ed. Wil Roebroeks and Thijsvan Kolfschoten, 297–315. Leiden: University of Leiden.

Rolland, Nicolas. 1981. The interpretation of Middle Paleo-lithic variability. Man 16:15–42.

———. 1988. Observations on some Middle Paleolithic timeseries in southern France. In Upper Pleistocene prehistory


of western Eurasia, ed. H. Dibble and A. Montet-White,161–80. University Museum, University of Pennsylvania,University Monograph 54.

Rolland, Nicolas, and Harold L. Dibble. 1990. A new synthesisof Middle Paleolithic variability. American Antiquity 55:480–99.

Ronen, Avraham, ed. 1982a. The transition from Lower toMiddle Paleolithic and the origin of modern man. BritishArchaeological Reports International Series 151.

———. 1982b. Concluding discussion. In The transition fromLower to Middle Paleolithic and the origin of modern man,ed. A. Ronen, 320–30. British Archaeological Reports In-ternational Series 151.

Sackett, James R. 1981. From de Mortillet to Bordes: A centuryof French Paleolithic research. In Towards a history of ar-chaeology, ed. G. Daniel, 85–99. London: Thames andHudson.

———. 1991. Straight archaeology French-style: The phy-logenetic paradigm in historic perspective. In Perspectiveson the past: Theoretical biases in Mediterranean hunter-gath-erer research, ed. G. A. Clark, 109–39. Philadelphia: Uni-versity of Pennsylvania Press.

Schick, Kathy D. 1994. The Movius Line reconsidered: Per-spectives on the earlier Paleolithic of eastern Asia. In In-tegrative paths to the past: Paleoanthropological advances inhonor of F. Clark Howell, ed. R. Ciochon and R. S. Cor-ruccini. Englewood Cliffs: Prentice-Hall.

Schick, K. B., and J. D. Clark. 2000. The Acheulean and thePlio-Pleistocene deposits of the Middle Awash Valley, Ethi-opia. In The Acheulean and the Plio-Pleistocene deposits ofthe Middle Awash Valley, Ethiopia, ed. J. de Heinzelin, J. D.Clark, K. B. Schick, and H. Gilbert, 123–81. Tervuren: Mu-see Royal de l’Afrique Centrale. [RGK]

Schreve, D. C. 2000. The vertebrate assemblage from Hoxne,Suffolk. In The Quaternary of Norfolk and Suffolk: Fieldguide, ed. S. G. Lewis, C. A. Whiteman, and R. C. Preece,155–63. London: Quaternary Research Association. [JM,HF]

———. 2001. Differentiation of the British Late Middle Pleis-tocene interglacials: The evidence from mammal biostra-tigraphy. Quaternary Science Reviews 20:1577–82. [JM, HF]

Schreve, D. C., D. R. Bridgland, P. Allen, J. J. Blackford, C.P. Gleed-Owen, H. I. Griffiths, D. H. Keen, and M. J. White.2002. Sedimentology, palaeontology, and archaeology ofLate Middle Pleistocene River Thames terrace deposits atPurfleet, Essex, UK. Quaternary Science Reviews 21:1423–64. [JM, HF]

Schvoerer, M., J. Rouanet, H. Navailles, and A. Debenath.1977. Datation absolue par thermoluminescence de restesantewurmiens de l’Abri Suard a La Chaise-de-Vouthon(Charente). Comptes Rendus de l’Academie des Sciences,Paris 284:1979–82.

Schwarcz, Henry, Bonnie Blackwell, and Andre Debenath.1983. Absolute dating of hominids and Paleolithic artefacts

of the cave of La Chaise de Vouthon (Charente), France.Journal of Archaeological Science 10:4–513.

Schwarcz, Henry, W. Buhay, Rainer Grun, Mary C. Stiner,Steven L. Kuhn, and G. H. Miller. 1990–91. Absolute datingof sites in coastal Lazio. Quaternaria Nova 1:51–68.

Schwarcz, Henry, and Andre Debenath. 1979. Datation ab-solue de restes humains de La Chaise-de-Vouthon (Cha-rente), au moyen du des-equilibre des series d’uranium.Compte Rendu de l’Academie des Sciences, Paris 288:1155–57.

Schwarcz, Henry, and Rainer Grun. 1993: Electron spin res-onance (ESR) dating of the Lower Industry. In The LowerPaleolithic Site at Hoxne, England, ed. R. Singer, B. Glad-felter, and J. Wymer. Chicago: University of Chicago Press.

Semaw, S., P. Renne, J. W. K. Harris, C. S. Feibel, R. L. Bernor,N. Fesseha, and K. Mowbray. 1997. 2.5-million-year-oldstone tools from Gona, Ethiopia. Nature 385:333–36.[RGK]

Singer, Ronald, Bruce G. Gladfelter, and John J. Wymer, eds.1993. The Lower Paleolithic site at Hoxne, England. Chicago:University of Chicago Press.

Singer, R., J. Wymer, B. G. Gladfelter, and R. G. Wolff. 1973.Excavation of the Clactonian industry at the Golf Course,Clacton-on-Sea, Essex Proceedings of the Prehistoric Society39:6–74. [JM, HF]

Soressi, Marie, D. Armand, F. d’Errico, H. L. Jones, E. Pubert,W. J. Rink, J.-P. Texier, and D. Vivent. 2002. Pech-de-l’AzeI (Carsac, Dordogne): Nouveaux travaux sur le Mousteriende tradition acheuleenne. Bulletin de la Societe PrehistoriqueFrancaise 99:1–7.

Straus, L. G. 1975. A study of the Solutrean in Vasco-Can-tabrian Spain. Ph.D.diss., University of Chicago. [LGS]

———. 1983. From Mousterian to Magdalenian: Cultural evo-lution viewed from Vasco-Cantabrian Spain and PyreneanFrance. British Archaeological Reports International Series164.

———. 1985. La collection Magnant: Un temoin de la pre-histoire charentaise. Bulletins et Memoires de la Societe Ar-cheologique et Historique de la Charente 1985(1):21–34.[LGS]

———. 1987. Paradigm lost: A personal view of the currentstate of Upper Paleolithic research. Helinium 27:157–71.[LGS]

———. 1991. Paradigm found? A research agenda for thestudy of the Upper and Post-Paleolithic in Southwest Eu-rope. In Perspectives on the past, ed. G. A. Clark, 56–78.Philadelphia: University of Pennsylvania Press. [LGS]

———. 2005. A mosaic of change: The Middle-Upper Pa-leolithic transition as viewed from New Mexico and Iberia.In Armageddon or entente? ed. L. G. Straus, 47–67. Oxford:Quaternary International 137/Elsevier. [LGS]

Straus, L. G., and G. A. Clark. 1986. La Riera Cave. Anthro-pological Research Papers 36. [LGS]

Stringer, Christopher, and J-J. Hublin. 1999. New age esti-


mates for the Swanscombe hominid and their significancefor human evolution. Journal of Human Evolution 37:873–77.

Taschini, M., and Amilcare Bietti. 1979. L’industrie lithiquede Grotta Guattari au Mont Circe (Latium): Definitionculturelle, typologique et chronologique du Pontinien.Quaternaria 21: 179–247.

Tattersall, Ian. 1995. The fossil trail. New York: Oxford Uni-versity Press.

Tixier, Jacques. 1984. Prehistoire de la pierre taillee. 2. Econ-omie du debitage laminaire. Paris: Cercle de Recherches etd’Etudes Prehistoriques.

Tixier, J., M.-L. Inizan, and H. Roche. 1980. Prehistoire de lapierre taillee. 1. Terminologie et technologie. Paris: Cercle deRecherches et d’Etudes Prehistoriques.

Tostevin, Gilbert B. 1996. Core reduction, chaıne operatoire,and other methods: The epistemologies of different ap-proaches to lithic analysis. Paper prepared for the AnnualMeeting of the Society for American Anthropology, SanJuan, P.R.

———. 2003a. A quest for antecedents: A comparison of theTerminal Middle Paleolithic and Early Upper Paleolithic ofthe Levant. In More than meets the eye: Studies on UpperPalaeolithic diversity in the Near East, ed. Nigel Goring-Morris and Anna Belfer-Cohen, 54–67. Oxford: OxbowPress.

———. 2003b. Attribute analysis of the lithic technologies ofStranska skala II-III in their regional and inter-regionalcontext. In Stranska skala: Origins of the Upper Paleolithicin the Brno Basin, ed. Jiri Svoboda and Ofer Bar-Yosef,77–118. Cambridge: Peabody Museum, Harvard Uni-versity.

Tozzi, C. 1970. La Grotta di S. Agostino (Gaeta). Rivista diScienze Preistoriche 25:3–87.

Trigger, B. G. 1989. A history of archaeological thought. Cam-bridge: Cambridge University Press.

Tryon, C. A., and S. McBrearty. 2002. Tephrostratigraphy andthe Acheulean to Middle Stone Age transition in the Kap-thurin Formation, Kenya. Journal of Human Evolution 42:211–35. [RGK]

Tuffreau, Alain. 1982. The transition Lower/Middle Palaeo-lithic in northern France. In The transition from Lower toMiddle Palaeolithic and the origin of modern man, ed. A.Ronen, 137–49. British Archaeological Reports Interna-tional Series 151.

———. 1996. Avant-propos. In L’Acheuleen dans l’Ouest del’Europe: Actes du Colloque de Saint-Riquier, ed. Alain Tuf-freau. Lille: Universite des Sciences et Technologies de Lille.

———. 2001a. Le gisement de Cagny-la-Garenne II(Somme): Introduction. In L’Acheuleen dans la vallee de laSomme: Donnees recentes, ed. A. Tuffreau. Lille: Universitedes Sciences et Technologies de Lille.

———. 2001b. Le gisement de Cagny-l’Epinette: Introduc-tion. In L. Acheuleen dans la vallee de la Somme: Donnees

recentes, ed. A. Tuffreau. Lille: Universite des Sciences etTechnologies de Lille.

Tuffreau, Alain, and Pierre Antoine. 1995. The earliest oc-cupation of Europe: Continental Northwestern Europe. InThe earliest occupation of Europe: Proceedings of the Euro-pean Science Foundation Workshop at Tautavel (France), ed.W. Roebroeks and T. van Kolfschoten, 147–64. Leiden: Uni-versity of Leiden.

Tuffreau, Alain, Pierre Antoine, Philip G. Chase, Harold L.Dibble, Brooks Ellwood, Thijs van Kolfschoten, Agnes La-motte, Michel Laurent, Shannon McPherron, Anne-MarieMoigne, and Andre V. Munaut. 1995. Le gisement acheu-leen de Cagny-l’Epinette (Somme). Bulletin de la SocietePrehistorique Francaise 92:169–91.

Tuffreau, Alain, Agnes Lamotte, and Jean-Luc Marcy. 1997.Land-use and site function in Acheulean complexes of theSomme Valley. World Archaeology 29:225–41.

Tuffreau, Alain, and Jean Somme. 1988. Le gisement paleoli-thique moyen de Biache-Saint-Vaast (Pas-de-Calais): Stra-tigraphie, environnement, etudes archeologiques. Memoiresde 1a Societe Prehistorique de France 21.

Utrilla, Pilar, Juan Vılchez, Lourdes Montes, Ignacio Baran-diaran, Jesus Altuna, Enrique Gil, and Pilar Lopez. 1987.La cueva de Pena Miel, Nieva de Cameros, La Rioja. Ex-cavaciones Arqueologicas en Espana 154.

Valladas, Helene, J. P. Chadelle, Jean-Michel Geneste, J. Joron,Lilianne Meignen, and Pierre-Jean Texier. 1987. Datationspar la thermoluminescence de gisements mousteriens duSud de la France. L’Anthropologie 91:211–26.

Valladas, Helene, Norbert Mercier, Jean Louis Joron, ShannonP. McPherron, Harold L. Dibble, and Michel Lenoir. 2003.TL dates for the Middle Paleolithic site of Combe-CapelleBas, France. Journal of Archaeological Science 30:144–50.

Valoch, Karel. 1982. The Lower/Middle Palaeolithic transitionin Czechoslovakia. In The transition from Lower to MiddlePalaeolithic and the origin of modern man, ed. A. Ronen,193–201. British Archaeological Reports International Se-ries 151.

Vega Toscano, L. G., L Raposo, and Manuel Santonja. 1999.Environments and settlement in the Middle Paleolithic ofthe Iberian peninsula. In The Middle Palaeolithic occupationof Europe, ed. W. Roebroeks and C. Gamble, 23–48. Leiden:University of Leiden.

Villa, Paola. 1991. Middle Pleistocene prehistory in south-western Europe: The state of our knowledge and ignorance.Journal of Anthropological Research 47:193–218.

Warren, S. Hazzledine. 1926. The classification of the LowerPaleolithic with special reference to Essex. Transactions ofthe Southeast Union of the Scientific Society 31:38–50.

White, M. J., and N. Ashton. 2003. Lower Palaeolithic coretechnology and the origins of the Levallois method innorth-western Europe. Current Anthropology 44:598–609.[JM, HF]


White, M. J., and S. Plunkett, eds. 2004. Miss Layard excavates:A Palaeolithic site at Foxhall Road, Ipswich, 1903–1905. Liv-erpool: Western Academic and Specialist Press. [JM, HF]

Wymer, John, and Ronald Singer. 1993. Flint industries andhuman activity. In The Lower Paleolithic site at Hoxne, En-gland, ed. R. Singer, B. Gladfelter, and J. Wymer, 74–128.Chicago: University of Chicago Press.

Wymer, John J., Bruce G. Gladfelter, and Ronald Singer. 1993.The industries at Hoxne and the Lower Paleolithic of Brit-ain. In The Lower Paleolithic site at Hoxne, England, ed. R.Singer, B. Gladfelter, and J. Wymer, 218–24. Chicago: Uni-versity of Chicago Press.

Wynn, T. 2002. Archaeology and cognitive evolution. Behav-ioral and Brain Sciences 25:389–438. [RGK]

Documents

The Lower/Middle Paleolithic Periodization in Western Europe